Although the emotional impact of media and technologies has been addressed since the mid-20th century, over the past two decades, the omnipresence of digital interfaces has generated everyday affective experiences that previous generations could not even imagine. Designed to elicit emotional responses, these interactions have a significant individual and social psychological impact, redefining consumption and governance on a global scale. Rather than being a comprehensive analysis, this text reflects on how art can function as a space to reveal, resist, or resignify the use of these interfaces and systems, rethinking what the aesthetic and artistic experience means in a world increasingly mediated by machine subjectivity.
Affective computing
When Rosalind Picard introduced the concept of affective computing in the 1990s, she did not anticipate the impact that emotion quantification would have on the design of digital interfaces. The MIT engineer emphasized the importance of emotions in human intelligence and argued that a purely logical approach was insufficient for creating more intuitive computer systems and with decision-making abilities. She proposed that to enhance human-machine interaction, computers should recognize, interpret, and respond to users’ emotions. This initiated a series of research focused on quantifiably analyzing emotional states through facial expressions, captured by cameras, as well as through physiological signals such as heart rhythm and skin conductivity, collected via biometric sensors.
Advances in this field have been remarkable, particularly regarding accuracy, thanks to the use of neural networks that redefine it as emotional artificial intelligence. However, two unclear issues persist in its implementation. The first relates to the bias and oversimplification of human emotions, such as those interpreted through facial gestures, which are often too simplistic compared to people’s real emotional spectrum, especially in neurodivergent individuals, for instance. The second problem pertains to the implications of using these interpretations to constantly and extensively monitor people’s subjectivity.
Affective tokenization and normalization of subjectivity in machines
Although affective computing has been utilised in education and health to design learning experiences, provide support, and ensure emotional follow-up, in other areas, emotions have become a tradable asset. Each interaction with the interfaces is analysed, and their results, in affective terms, are tokenized. Social media, for instance, favours the viral spread of content that generates extreme emotional responses, which has political implications, as polarizing speeches garner more attention than critical analysis. In this context, governance is influenced by the manipulation and mass exploitation of emotions through digital media.
And this information overload significantly impacts us not only emotionally but also psychologically. Our brains are constantly bombarded with data that exceeds their processing capacity. Prolonged exposure to distressing images and news can lead to desensitization, diminishing empathy and fostering a sense of helplessness towards global issues, contributing to social isolation and a focus on consumerism. However, rather than questioning information or its sources, or simply stepping back from the overstimulation, the trend seems to lean towards seeking answers within the same technology. ChatGPT Jailbreaks, like Dan, have gained immense popularity among those who have given up face-to-face relationships, opting instead for platonic connections with virtual partners, claiming that these interactions bring greater happiness than contact with real people. An increasing number of individuals are also choosing chatbots over engaging in therapy with human psychologists.
The artistic experience in the age of AI
From the perspective of artistic production, the relationship between art appreciation and sensitive emotional experience has been discussed for centuries. With Romanticism, emotion played a central role in the aesthetic experience, while in the 20th century, artistic avant-gardes challenged the boundaries of perception and affectivity in the artistic process. Today, in a context of digitization and the proliferation of artificial intelligence, these notions are being challenged. Faced with an audience saturated with stimuli, the question arises whether creations generated by artificial intelligence can elicit emotional responses comparable to those of human art. Many artists express concerns about reaching an emotionally saturated audience. However, artistic production operates on multiple levels involving observation, interpretation, and complex contextualization, which remain unique to human perception and judgment.
A fundamental part of appreciating and enjoying art lies not only in the emotion it sparks, but also in its conceptual dimension and the experience of discovery within the framework that defines artistic production in this century. The growing influence of technology on contemporary thinking is reflected not only in the incorporation of electronics into art installations but also in a deeper transformation: the conception of artwork as a system. This perspective implies that the work does not operate in isolation but is rather situated within the dynamics of other systems with which the public interacts. In this regard, Benjamin Bratton, in The Stack: On Software and Sovereignty (2016), proposes a framework to understand how digital infrastructures reconfigure our relationships with the environment. This idea resonates with emerging logics in today’s art practice.
Art and its technological mediations
Regarding the relationship between emotions and artistic production and appreciation, we are facing an unprecedented moment in history: for the first time, there are tools designed to quantify emotional responses to works of art. However, this measurement does not determine the significance or value of the work; instead, it opens a new field of discussion. The ability to register emotion quantification at the centre of the artistic experience transforms not only the perception of the work but also the interaction between viewers and their relationship with space.
An example of this exploration is Lauren Lee McCarthy’s Vibe Check, where audience faces are analysed in real time, translating their emotional responses into images and text. The experience, simultaneously playful and uncomfortable, shifts the focus of the work from an object of contemplation to the emotional reaction of the viewer observing others who share the same space, generating a game of self-observation and collective perception.
If emotions have historically been a central axis in the relationship with art, to what extent does the ability to quantify and manipulate them alter that relationship? can art, precisely because of its intrinsic bond with affection, offer a reorientation or resistance to the emotional instrumentalization that prevails in digital media? And if both art and media predispose us to an emotional experience – whether consciously or unconsciously – what differentiates aesthetic emotion from that arising from the interaction with algorithmic environments designed to model and predict our responses?
This approach opens the debate about technologies and their impact on aesthetic appreciation. It also challenges whether art, in its relationship with sensitiveness, can function as an agency space against the logics of automation and emotional surveillance.
About interfaces
Before exploring the possibilities of affective computation and machine subjectivity in art –including the creation of systems that respond to emotions – it is crucial to consider how the artist’s relationship with technology impacts the discourse of the work. I’ve noticed that many young artists approach certain digital tools as if they were black boxes, fascinated by their ability to generate amazing results from minimal inputs, as is often the case with AI-powered video and image production. However, using closed technologies for business purposes imposes limits and enrolls them in consumer and surveillance logic. This does not prevent their use in art but requires a critical look that reveals their nature and problematizes their impact.
Further exploration involves hacking these systems, i.e., their intervention to deflect their purpose and generate unexpected or even contrary results to their original programming. The artist’s optimal relationship with the technology they employ is not limited to conceptual understanding; it also encompasses the technical knowledge necessary to design their own systems, either individually or in collaboration. This approach more closely resembles architecture design, not only in problem-solving but also in anticipating how the public will interact with, intervene in, or be intervened by the work. Thus, the interfaces that the artist produces are not the conventional ones found in commercial spaces but rather artefacts that establish unlikely connections between objects, software, bodies, and human presences, articulating a transmedia machinery with a speech of its own.
Technoaffective art
From the various conceptual approaches of artists to affective computing and their dialogue with digital subjectivities, we can outline a catalogue of works for subsequent analysis. The first approach is the critical exposure of how emotion quantification is integrated into surveillance and control systems. Trevor Paglen’s work exemplifies this perspective, utilizing digital surveillance and artificial intelligence tools to connect documentation, databases, images, sculpture, and real-time video. His works demonstrate how technologies such as facial recognition and imaging operate as mechanisms of ideological manipulation. An example of his work, which is close to emotional recognition, is The Trolls (2019), where he prints a database used to train an artificial intelligence capable of detecting offensive or abusive language on social media.
Some works serve as interfaces for exploration and emotional reflection for participants. These often playful experiences invite audiences to interact with their own bodies and use their subjectivity to generate different reactions in the work. Examples of this are Mood Swings (2009), from Bialoskorski, Westerink, and Broek, where the system interprets body movements as affective expressions and uses a colour and light model to feed back the experience, or my own piece Environmental Disturbances (2012), where the public had access to a brainwave sensor with which a landscape narrative designed to induce a state of calm was created. Both proposals explore biofeedback through affective computing, offering the public an avenue for introspection and perhaps brief moments of emotional connection.
Among the works that question the relationship between technology, subjectivity, and emotions, some explore the interstitial spaces of machine subjectivity, generating more questions about its future than definitive answers regarding human, social, or political issues. An example is Mario Klingemann’s Memories of Passersby I (2019), where portraits generated by artificial intelligence evoke gestures and expressions with an unsettling familiarity, but lack a defined human identity. Another case is The Blind Robot (2015) by Louis-Philippe Demers, a work in which a robot explores the faces of viewers by touch while they are reflected in a mirror, confronting the strangeness of being emotionally read by a machine without vision.
In human-machine interaction, affective computing shapes our perception of reality by interpreting and responding to emotions within technological frameworks geared towards consumption, surveillance, and prediction. By translating subjectivity into data, these systems reshape our understanding of feeling. However, art serves as a space of resistance where these dynamics can be evidenced and questioned. It is possible to reveal its scopes and limits through experimentation with affective computing, exposing the algorithmic mediation of emotions. Thus, art not only confronts systems that seek to define us, but also rethinks what it means to be human in a world traversed by artificial intelligences that attempt to decipher and shape our emotions.
Recommended citation: GARZA LAU, Anni. AI, my love. About the subjectivity of machines and the artistic experience. Mosaic [online], May 2025, no. 203. ISSN: 1696-3296. DOI: https://doi.org/10.7238/m.n203.2506
Deja un comentario