Capítulo / Chapter III | Cinema – Comunicação / Communication

The Cinematic Selfie: Questioning the Self through Generative Art

Pedro Alves da Veiga

Centro de Investigação em Artes e Comunicação, Universidade Aberta


This article addresses and questions the magic-mirror phenomenon, popularised by current smartphone selfie and video capture apps. This phenomenon stimulates the illusion of control over the appearance of the face, either through applying semi-automatic soft filters to highlight the face area, to smooth the skin or correct the posture; or through the use of humorous add-ons or distortions, such as bunny ears or anime features, among others. However these results are short-lived, as their publication in social networks is either ephemeral – as a story – or timed to become invisible or irrelevant – in the timeline stream. Cumulatively they leave little margin (if at all) to stimulate a deeper reflection on the subject of (self) identity, and could thus be reduced to an expression of narcissism and consumption rather than a shared, transformative, meaningful practice. The two generative artworks described in this article, on the other hand, seek to guide the visitor beyond the visual magic-mirror through thought-provoking and reflective processes, where face-based audio-visual trance inducing cycles are used to hint at new identities and possibilities, challenging species, race, gender and age. These artworks seek to immerse the visitor, with narrowed awareness of external surroundings and stimuli, with a deepened focus in a synesthetic experience of flow, aiming at an altered perception of the self. If appearance can act as a tool to communicate one’s identity to others, this article ponders the possibility that such a synesthetic environment can be artivistically used to influence the perception of the self.

Keywords: Artivism, Generative art, Interactive art, Cinematic art


When MySpace, Flickr, LinkedIn and Facebook first started to gain traction at the dawn of Web 2.0, the key-concepts of social network, social media, participation or user-generated content sprung to the spotlight. A haven of creativity and their blossoming potential for culture, art, science, creativity, philosophy, experimentation, exploration and adventure was heralded and likened to a reboot of humanity, anchored on the individual and their dialectical tension with the collective. In 2006 Time magazine acknowledged “you” as the person of the year, its cover bearing a photograph of a computer whose screen had been replaced by a mirrored surface. This individual, at the centre of all attentions, was described by Twenge (2013) as Generation Me and characterised as overconfident, with high expectations, higher narcissism, lower creativity, are less interested in civic issues. Generation Me has been shaped by decades of audio-visual streaming and television watching, as documented by the quasi-anecdotal CSI effect (Hawkins and Scherr, 2017), acquiring pseudo-expertise in all areas. However, unlike the CSI specialised scientific equipment, other real tools (for cooking, landscaping, interior design, music making, creative writing, visual arts, etc.) became easily and readily available on the market – some would say democratised while others would prefer the term massified – in a technological explosion of devices, heir to the desktop publishing revolution of the 1980’s, the home video cameras of the 1970’s and the family stills cameras of the 1950’s. At that time social networks, newspapers and television networks started asking readers and viewers to submit their own content (Veiga, Tavares and Alvelos 2017).

Out of the possible motivations for such behaviours, two of them seem to take the lead: money and reputation (Origgi 2019), and since these contributions are mostly unpaid, it seems that reputation is the key to unpaid spontaneous collaboration in a society where exposure is synonymous with success (Markos, Labrecque and Milne 2012).

The exposure of the individual has thus acquired a similar status to that of socialites and entrepreneurs who rise to stardom. The rise of celebrity culture has not happened by itself: it has been cultivated by agencies, producers and the media. And it has a function: the more distant and impersonal the multinational giants are, the more they depend on familiar faces to instil confidence and establish a personalized connection – whose business fallacy is not always obvious – with their customers. If beforehand this role was entrusted to movie and music stars, today it is the Internet star that takes charge, acting from the local to the international scale.

We conceptualize celebrity as an organic and ever-changing performative practice rather than a set of intrinsic personal characteristics or external labels. This practice involves ongoing maintenance of a fan base, performed intimacy, authenticity and access, and construction of a consumable persona. (...) On Twitter, performative intimacy is practiced by posting personal pictures and videos, addressing rumors, and sharing personal information. Picture-hosting services, such as YFrog and Twitpic, allow users directly to post cameraphone pictures to Twitter. Famous people frequently use these services, creating the illusion of first-person glimpses into their lives. Ashton Kutcher, for example, tweets pictures of himself on set, during talk-show appearances, and posing with his wife Demi Moore and celebrities such as actress Mischa Barton and R&B singer Usher. (Marwick and Boyd 2011, 140 and 148)

Young people in particular appear to be more obsessed with the celebrity status and fame, which is mostly built around the projection of a carefully crafted image of the self, and research does suggest that narcissism is on the rise, with each successive generation surpassing the levels of the previous one (Young et al. 2016). Media content aimed at pre-teens in the 21st century exerts a greater appeal to individualism and sensationalism than in the 1990s, facilitating a narcissistic pre-teen culture (Rosen 2016).

The relentlessly repeated catch-phrases and affirmations in sales pitches, slogans, personal motivation manuals and lectures on creativity shape public opinion: the creative individual is idolized and advertised as being hyper-empowered, directed towards fame, using myriads of technological tools, moving between the physical and virtual universes.

Reality TV catapults complete strangers to fame, based solely on the networks’ own aesthetic and audience building criteria, leading to the replaceable celebrity-commodity coined by Turner (2006: 161). The perception that anyone, seemingly ordinary and similar to so many others, can become a media star is one of the likely causes of the widespread hunger for fame: a survey of 16-year-olds in the UK revealed that 54% of them dream of becoming celebrities1.

Fame is mostly coveted by the individual, even if supported by larger (infra)structures. Famous artistic collectives, whose contemporary materializations in the field of music consist in bands – especially boy bands and girl bands – are systematically dismembered after a while due to the individual pursuits of this status, which seems to indicate that the phenomenon of individualism runs parallel to that of fame.

As a consequence of the introduction of new and accessible technologies in the market, everyone can easily create videos, upload photographs and engage with other users on a plethora of subjects – including the self. In the globally aestheticised and exposure-addicted society (Lipovetsky and Serroy 2016) there is a massive production of aestheticised audio-visual content, reflecting the ease of access to technology but also the need to use it for individual expression. Under the influence of a global consumer market, individuals have acquired multiple skills, not so much in the romantic way that idealizes the multidisciplinary (re)Renaissance man, but because they feel the need to rise above the noise floor of the other millions who are behaving in a similar way, attracting more followers and seeking the much coveted viral effect. Thus they had to acquire some of the characteristics of entrepreneurs, marketers and journalists – and the influencer was born. Gaining more exposure – as a measure of success – allows influencers to claim a status that will allow them to get paid for the number of views – not necessarily the quality of the content they promote – using marketing and communication techniques, many of them more complex than their actual publications. The services and tools that help these influencers – from self-help manuals to financial services, website hosting, online ad management and placement, software and creative toolkits, used to design, produce, present, communicate, market, exhibit and sell their work, services and products – are among the most successful businesses of our time.

There is, however, one common trait: the personal image. All influencers present themselves in carefully crafted poses, looks and settings. In the neoliberal society individuals are expected to be creative and innovative – like artists – but also to lead creative and innovative lives, visiting spectacular and aesthetically pleasing sites – and share its evidence. One could then conclude that, if anything, influencers are indeed influencing a culture of self-image and self-promotion, where buzzwords such as creativity and innovation abound.

But what if creativity is only a business concept, in line with other clichés, such as leadership, service disruption, innovation, and transformation, as posited by Deresiewicz (2015), and the bombardment of concepts built around the self is a part of this strategy? Self-help, self-improvement, self-respect, self-recognition seem to always imply acquiring something, be it a book, an online course or the services of a life-coach, all leading to the ultimate evidence of daily accomplishment: the selfie2.

Contemporary uses of media involve an increasing use of digital(ized) faces, used as profile photos, avatars and stickers, combining an unprecedented quantity and quality of images in new forms of interaction. The emergence of face-based cultural practices can be attributed both to technological innovations (digital photography and filters, software for automatic face recognition), and to the dissemination of different face representations, such as the selfie, animojis, stickers and memes.

Digital Narcissus

Back in 2013 Marche deemed the selfie so dominant a form that it might even be shaping technology (Marche 2013). And in fact many products have been developed in order to make better, easier selfies, from the selfie stick, to the selfie circular USB lighting for smartphones.

Face detection and recognition are among the social skills that are acquired very early in human life. They are key elements of social interaction as they help ascertain socially meaningful information in terms of familiarity, attractiveness, and emotional status, which will then shape behavioural patterns (Lopatina et al. 2018).

It is probably one of the oldest human traits to be fascinated by one’s own reflection. The Roman poet Ovid told the story of Narcissus, a young hunter known for his beauty. Narcissus rejected the romantic advances of the nymph Echo and so the Goddess Nemesis enticed him to a pool of water where Narcissus was so mesmerized by his own reflection that he eventually perished by the waterside, unable to break away from gazing at his own image.

This fascination is also expressed in artistic renderings, especially portraits and self-portraits, and their modern versions, such as the very large amount of mirrored surfaces surrounding us and in the technology that enables us to capture such reflections, mostly embodied in the ever-present selfie: on the 28th June 2021 a quick search for the hashtag #selfie on Instagram produced over 450 million results.

The term narcissism, after the myth of Narcissus, has been used to denote fascination with one’s physical beauty as well as general and unrestrained forms of self-love. Freud most significantly brought narcissism under the lens of psychology, and proposed that narcissists publicly portray an improved and perfected image of themselves, oftentimes as a form of compensation of low self-esteem and self-worth (March and McBean 2018).

Thus, the public display of one’s own image or portrait implies extra care in choosing the most favourable profile side, hair styling, and some individuals may even go as far as always using make-up products, either physical or virtual, with popular software tools like Photoshop and other smartphone apps, such as Instagram, Snapchat and FaceApp ranking among the most widely used.

The “self” represented in the photo is clearly understood as carefully crafted and aestheticized, and presenting visually in a particular moment (i.e., in this particular light, from this angle, with this arrangement of facial features, body parts) as afforded through a particular technological arrangement (i.e., with this light filters, editing apps, or camera). Digital photography allows for this careful parsing out of instances and potentialities, producing many possible variations of the self’s image and the availability of the face for scrutiny: manifested through time and expertise, emphasizing the ephemeral, and lived experience as much as it does calculation. (Lavrence and Cambre 2020, 7)

Casey suggests that the rise of the smartphone camera will turn out to be as important to the history of art as the appearance of the mirror, which led to new ways of understanding the self through art (Casey 2017). Other critics may suggest that the selfie-culture has fostered narcissism to unprecedented heights, but some scholars have argued that this practice is actually empowering, particularly for populations who have historically been denied access to public self-representation.

There are famous, innocuous and infamous selfies3, as well as equally questionable collections of thematic selfies, such as the ones shared in the Facebook group “With My Besties in Auschwitz”4, the “Selfies at Funerals” tumblr5, or even the Killfie6 and Belfie7 trends. But the vast majority of selfies are ordinary, easily forgettable and there is no indication or evidence that they might foster any type of self-analysis or reflection beyond their immediate social value or self-esteem compensation mechanisms (March and McBean 2018).

Animated selfies were introduced by several apps that changed their algorithms, like Vine (to take six-second video selfies), Instagram (which added video support to allow for 15-second video selfies) and Snapchat (allowing for potentially embarrassing or even x-rated selfies, without worrying about them lingering on the Web), and over the years these animated forms have also become increasingly popular in other smartphone apps such as Tinder or TikTok. This return to the animated format, capturing those moments where the individual is watching their representation on the screen of a mobile application, usually on a smartphone or tablet, finds its roots in the history of media technology, where the video camera – one of the first devices to offer the simultaneous technological image – and closed-circuit video can be regarded as a cultural and technological predecessor to the selfie.

An existential space can either become stratified and caught in repetitions – as is the case of television viewers who sit in front of theirs screens binge watching series and films, or the Instagrammers who produce different versions of the same photographic compositions8 – or has the potential to lead to new dimensions and foster new knowledge. Guattari (2000) described such a catalysing power as micro-political: a system has a certain amount of potential energy, which can be used to produce significant transformation. However it requires a particular configuration or an energy spark in order to enter that state of transformation. If this singular event does not occur, the system will thus remain in the same unchanged state.

How would the modern selfie-addicted audience react to an artistic cinematic rendition of their own faces, stepping out of their aesthetic control while still clearly allowing for their identification/recognition? How would they feel if suddenly they would gain traits of a different gender, age, race or culture? Would they strive to look for the self in the image or would they reject it? Would it bring out their playful nature or would it cause discomfort and scare them away? And what role would other stimuli – like audio and text – play in altering their perception of such images and overall engagement?

Modulating the Selfie

The term modulate derives from the Latin modulari, which is itself derived from modus, meaning bound, limit, manner, mode, way, method, rule, rhythm, beat, measure, or size. In its original use, related to music, it is a technical term used to describe a change of key or volume during execution. By similarity it has come to denote a similar alteration in other areas, implying a regulated change into a different register, condition or form by exerting a modifying or controlling influence.

The author currently posits its use aligned with all of the above meanings, in a manner not dissimilar to Simondon’s usage of the term, according to Combes (2013, 15), which is “the putting into relation of an operation and a structure”. The act of modulating a source signal implies a change to the signal itself, but also to its perception and the impacts it may cause upon the audience.

(...) the principle of individuation is the operation that carries out an energy exchange between the matter and the form, until the unity leads to a state of equilibrium. One could say that the principle of individuation is the common allagmatic operation of the matter and form through the actualization of potential energy. This energy is energy of a system; it can produce effects in all the points of the system in an equal way, it is available and is communicated. This operation rests on the singularity or the singularities of the concrete here and now; it envelops them and amplifies them. (Simondon 1964, 44)

Combes posits that the principle of individuation is, in fact, modulation (2013: 5), as Simondon’s understanding of the term covers all nuances and fields where it is relevant. He assigns equal weight to all participating agents in the act of modulation, each one influencing the other, since the modulation process can thus be understood as occurring in different – even opposing – directions. For Simondon (1992), information is not just conserved as it is magnified and amplified in the modulation process that individuates a new system, structured from the previously disparate agents, but where new informational structures and meanings are created and governed. The experience of modulation is actually a process. When it comes to digital images, Groys describes this experience as a performance:

Similarly, looking at digital images we are also confronted every time with a new event of visualization of invisible data. So we can say: The digital image is a copy—but the event of its visualization is an original event, because the digital copy is a copy that has no visible original. That further means: A digital image, to be seen, should not be merely exhibited but staged, performed. Here the image begins to function analogously to a piece of music, whose score, as is generally known, is not identical to the musical piece—the score itself being silent. For music to resound, it has to be performed. Thus one can say that digitalization turns the visual arts into a performing art. But to perform something is to interpret it, to betray it, to distort it. Every performance is an interpretation and every interpretation is a betrayal, a misuse. (Groys 2008, 84)

When digital data is modulated into a sequence of pixels and displayed as an image, every individual instance of its experience is as different from one another as the same musical score played by different orchestras in different concert rooms in different occasions. But this modulation can even prevent the predictability of the experience: as random factors may come into play and the image itself is not static, either because it is being captured as a video stream or because it is being disassembled and reassembled in the process, as is typical with generative art, privileging a dynamic process over a static result.

The increasing use of faces in digital media culture has also fostered the parallel pervasiveness of animated anthropomorphic faces, derived from previous popular culture icons such as Mickey Mouse, Hello Kitty, the Linux penguin along with a plethora of mascots. This modulation of the human with non-human characteristics is well rooted in the past, as can be observed in several Egyptian deities and mythological creatures. These modulated uses of the selfie can enable communication between the inner and outer layers of the cultural self-description systems (identity–alterity), linking the somatic and the semiotic, the natural and the cultural, and act as a mediatory interface with the heterotopian (animist/totemist) ontology. The presence of and interaction with heterotopic characters could be linked to the inception of new semiotic codes and models of behaviour (Lotman 1990, 233–234).

Generative art

The designation generative applied to art was first introduced by Georg Nees in 1965, with his Stuttgart exhibition “Generative Computergraphik”. There are various definitions and classifications of generative art (Galanter, 2014; McCormack et al., 2014) as well as alternative designations for various subtypes, such as systems art, interactive art, algorithmic art, OpArt, BioArt, evolutionary art, among others. However, the term generative always implies the existence of an autonomous algorithmic structure that leads to the creation of a certain output – and the artwork consists in the ensuing runtime process and its output, rather than its encoding. The algorithmic structure is used to combine order (specific rules) with chaos (controlled randomness, interaction), each iteration becoming the seed for the next one, thus resulting in a seemingly infinite sequence of states – a flow –, but all within a certain aesthetic boundary defined by the artist/programmer. These systems also may vary in terms of their sensitivity to initial and external conditions, and can be defined as non-sensitive (closed) or sensitive (open). Non-sensitive systems produce a finite number of states, even if in a very large number that will probably prevent repetition to occur during the audience experience of the artwork. In such systems the end result is defined by the algorithmic structure alone and has no significant dependency on the initial generation or on external factors, thus closed systems are usually non-interactive or with limited interactivity. Sensitive systems, on the other hand, will eventually generate a potentially infinite number of elements, not only because the state strongly influences its evolution, but also because external interference and interactivity is allowed (and welcome) and they contribute to further modulate the end result (Veiga 2017).

Generative systems have also been used in various degrees of relationship with the human face: either as Generative Adversarial Networks for generating artificial – yet photorealistic – faces (Moura and Ferreira-Lopes 2017), for facial anonymization (Hukkelås, Mester and 2019) or face restoration (Wang, et al. 2021), among others, thus deeming the generative approach as suitable and current to the topic in analysis.


Alchimia is an open generative artwork, and was designed as a magic-mirror of sorts. Far from fostering narcissism, Alchimia aims at provoking the selfie-addicted audience with renderings of their own faces, modulated out of their aesthetic control, yet delivering easily identifiable portraits and reacting to real-time changes in pose, expression or attitude. The artwork detects and captures a front facing face – and the face alone – and modulates each captured frame with different traits of other faces that were previously loaded into a library, containing a wide range of skin types, ages, gender, including tribal masks from around the world.

In an age where racism, sexual discrimination and beauty standards are still the subject of so much controversy, fostered by the ubiquitous pressure to look good and to post attractive photos (Gill 2021), Alchimia aims at questioning identity by continuously transforming each detected face, and bending gender, age and ethnicity over a continuous cinematic audio-visual flow.

There are two critical differences from the popular smartphone applications, whose filters are permanently under user control: (1) the complete lack of control the viewers can exert – they only control their own expression and pose, (2) the induction of an hypnotic, trance-like state with the help of an equally real-time generative soundtrack, which helps build a sense of climax and revelation, culminating in an oracle-like (teaser) text message being displayed. After this culmination the system starts afresh for a whole new cycle. Both these aspects define the synesthetic experience created by the flow of modulation and make it unique, not just for each viewer, but also for each of their experience-cycles with the artwork.

Figure 1 - A screen capture of Alchimia showing the original (pre-modulation) captured face in the upper right corner

Figure 2 - A screen capture of Alchimia where the original face is modulated by an Egyptian mask

In order to understand and document the behaviour of the interactors with the artwork, the author observed how and for how long different audience members were involved. He then engaged in qualitative interviews in order to gather insights, opinions and criticism. This data acquisition process has been continuous since the first public exhibitions of Alchimia, in its earlier versions, while still in an academic environment, in 2016, and to this day extends to a universe of 82 individuals, having travelled between different art venues in Portugal, Spain and China. These interviews determined several modifications in the algorithmic structure, and in its current state the artwork is faster and less serene than the initial version.

The teasers served two distinct purposes: they allowed for the current interactor to achieve closure in their interaction with the artwork, but they also were found to trigger the desire in other watchers (potential interactors) to take part in the performance, and thus present them with the opportunity to do so, replacing the current interactor. The numerous textual teasers include statements such as am I human?, is this me?, we see your soul, look into our eyes, you are not digital, come closer, is that your fear?, step into the light, reveal yourself, I cannot feel you, which are all open to interpretation as well as mystery. The interaction cycle length was adjusted to facilitate access to this moment of revelation. The sense of crescendo toward a climax was enhanced by means of the soundtrack, as well as a faster modulation of the face image, so that the interactor had less time to adjust to each new state. It was documented that some audience members remained for several cycles (the largest number of documented consecutive cycles was nine, with a rough temporal equivalence to nine minutes).

Figure 3 - A screen capture of Alchimia depicting one of the oracular teasers

Some observations were empirical and unplanned. As such, it was observed that the existence of other viewers tended to affect the behaviour and time spent in the experience, in line with previous findings (Frey 1978; Carver 2012). When alone the interactor would tend to spend more time engaging with the artefact, assuming a more immersed demeanour and disregarding their projected social image – poking tongue, eyes wide-open, funny faces – which consisted in the exact opposite of the carefully constructed selfie pose.

Most interactors conceded that the experience was hypnotic, and they felt as though they reached (mildly) altered states of consciousness, loosing track of time and place, always being anchored in their own image but letting the on-going modulation transport them elsewhere. However, as soon as an external presence was perceived, the exclusive attention dedicated to the experience would cease, and the dedicated time to the experience would also decrease. If this external presence were familiar, then engagement with the artwork would still be considerable, but almost reduced to strict entertainment, without any visible introspection or self-liberation. Some visitors admitted to engaging with the artwork only “to see what it tells me” (the oracle effect), and their close friends or relatives would subsequently dispute their place for the exact same reason (“let me try now”).

A direct screen capture of a real-time session can be seen at, as well as live footage documenting an audience member’s interaction, culminating at the textual teaser. This live footage clip was inserted in the original non-edited capture at approximately 2’34” through to 3’16”, where the unedited capture is resumed.

Due to the type of modulation and the broad spectrum of outcomes it is safe to assume that there actually is no predefined intention or meaning built into the aesthetic cinematic experience of the artwork. Instead there is a range of possible outcomes and experiences, also modulated by the cultural and existential inclination of each viewer. The direct relationship the artwork establishes with its interactor by means of the webcam and screen/projection creates an individual interactive-without-touch space/experience, which could almost be compared to a magic mirror, where total freedom of expression is welcome, as a means to assert (“that is actually me!”) or question (“is that really me?”) identity; but by using a large-size projection it can also be shared with other viewers, for whom the real artwork/performance is the actual interaction taking place between artefact and interactor.

Speciesism | Ageism | Racism

Speciesism | Ageism | Racism (SAR) is a closed generative cinematic artwork, using remixed photographic images and sound samples of familiar elements (photographic images of people and animal faces, human voices and animal sounds, musical instruments, natural atmospheric sounds, as well as synthetic sounds), algorithmically composed and modulated, and presented in an also familiar format, that of a cinematic audio-visual non-interactive continuous stream.

SAR was developed over two distinct conceptual vectors: an ontological perspective, questioning the righteousness of any demographic group to apply morality, concede rights and attribute worth only within the group itself, denying it to all outsiders and furthermore justifying the exploitation of those same outsiders for the benefit of the group, grounded upon an adversarial vision of the world, running contrary to an understanding of the world as a rhizome of interdependencies, where interconnected, collaborative, mutually dependent entities thrive; and a second aesthetic perspective, inspired by the worldwide mask-making tradition, rooted in animism, totemism, shamanism, ritual and mythology, stressing the relationship between human and nature, where the ideals of a permeating energy that binds every biological and non-biological entity reinforces the interconnectedness and collaborative nature of the world. The dual nature of the mask, perceived as both an exotic enigma and a familiar presence, implies that they are not mere pictures of other beings; they constitute a means to attribute or predicate the identity of those beings to the mask-wearer, to articulate power between the wearer and the seer and to appeal to psychological and cognitive processes. Masks have the function of transforming identity, either by modulating the representation of identity, or through its temporary and representational extinction and replacement with a diverse identity. They embody a modulation of the self(ie), through the transformation of the human into a being of another order (Pollock 1995).

Figure 4 - A combination of three different faces produces a new digital mask

SAR generates digital masks to represent and question various types of discrimination in modern times, merging ages, genders, races and species into one new entity, transforming the many into one new totemic ontology, which is characterized by a fusion of interiority and exteriority, a familiar process of hybridization, from antiquity to current popular culture, which undermines the encyclopaedic boundaries between species, races, ages and genders, as well as natural, artificial or cultural. Within this system, the digital masks can act as a double-layer mediatory interface, as they are generated from faces mostly donated by a participating audience, through an open and on-going call. Up to this moment a total of 63 contributions (selfies) from different audience members, sometimes including their pets’ faces, from eleven different countries across three continents were received, and are continuously incorporated into the artwork’s database. Human and animal faces are thus iteratively and generatively integrated into all exhibitions of the artefact. Each donated face is carefully pre-processed and split into three horizontal rectangles: eyes, nose and mouth. These slices are then recombined in a mask-totem composition, as it uses vertically stacked different beings, clearly identifiable as individual elements, and (vertically) symmetrical in composition.

The slices used to produce each mask are also further modulated into a Petri dish inspired setting where various samples of the elements being used are displayed around them.

Figure 5 - A screen capture of the continuous mask-generation process, depicting the “Petri dish” experimental environment around the central mask

Figure 6 - The “life-death” culmination of a full cycle in SAR

As with Alchimia, the progression of the generative process also increases its dynamics, until it reaches a new state. Unlike the oracle in Alchimia, this new state is the life-death cycle, common to all beings, human and non-human alike. At this point the screen is filled with words describing body parts and actions performed by living organisms and the system begins to slow down in pace and intensity. Finally death marks the end of the cycle, only for the whole generative process to start anew. Also like Alchimia, SAR uses a real-time generated soundtrack to achieve synchronicity with the visual animation, thus reinforcing SAR’s cinematic experience. This soundtrack relies on ambient noises of wind, rain, insects, birds, and also on distinct tribal rhythms and instruments from all over the world, modulated with digital sound synthesis and effects. Over the soundtrack different human voices utter “I am” in various languages, recorded through Google Translate voice synthesis function.

SAR can be seen at in its on-line pre-calculated version.

Twenty audience members, whose faces currently integrate SAR, agreed to provide feedback during screenings of the artwork. They were asked to describe the artwork in just one word/term, and so far twelve different terms were registered: beautiful (3), weird (3), funny (2), nice (2), amazing (1), intriguing (1) colourful (1), magic (1), disappointing (1), thought-provoking (1), interesting (1) and hideous (1).

Figure 7 - One of the mask combinations that sparked repulse, mostly due to recognizing part of the donated face mixed with a “less deserving” animal species

When asked if the use of different species, ages and races to generate a new entity was perceived as shocking, audience members stated they did not like one or more of the species combinations. One audience member felt the artwork failed to explicitly address the subjects of Speciesism, Racism and Ageism expressed disappointment. No one felt shocked by the remixing of human ages, sexes and races, but some expressed discomfort and dislike when the chimeras spanned across species, and two audience members dismissed speciesism as a “vegan thing” (sic).

It could be argued that, since there is no wearer, SAR does not actually generate masks, and is actually closer to a digital (dynamic) totem. However the author’s counter-arguments that SAR’s masks are worn at a virtual, psychological level, particularly by its participatory audience members, who seek to recognize their own participation (the eyes being the most commonly recognizable feature). As they gaze upon the generated images they simultaneously project the remixed entity onto themselves, identifying each animal part as being cast upon themselves – hence the rejection, in the above-mentioned cases. This was established during the informal interviews, particularly when the identification of the self occurred in a new unappealing entity. In one situation the observer was remixed with insect parts, and in the other with cephalopod parts, thus causing heightened aversion.


Generative cinematic artworks are not limited to the generation of abstract sequences, and may be successfully used together with figurative renderings to produce aesthetically pleasing results, as well as engage in artivism, beyond mere entertainment. Their potential for challenging preconceived notions is significant as is their potential for capturing the audience’s attention, therefore placing generative figurative art alongside other popular artivism art forms and types. In accordance with two studies conducted in museums (Smith, Smith and Tinio 2017), it was concluded that the average amount of time visitors spend engaging with artworks is below 30 seconds, thus for time dependent digital media artworks it is crucial to capture the viewers’ attention right from the first cycle of interaction and to keep them engaged, focused in reaching a specific reward or goal, as both artworks demonstrated.

Leaving room for interpretation is seen by the author as a positive trait in these artworks, allowing for a less biased analysis of visitor behaviour and interaction, particularly in situations where some of the visitors had no idea of the intention behind the works, their engagement was still remarkable, as were their assumptions and ideations. Some of these provided the author with ideas for future versions, such as linking the audience behaviour in terms of distance to the artwork, movement or demeanour to the modulation processes, in terms of speed, colour and element size, and also adapting the text messages in Alchimia to the above parameters. A new interactive version of SAR has been finalised allowing for faces to be dynamically captured and integrated into the generative flow.

There are different audiences and they will all have different expectations, but the new-media art audience may be regarded as more diverse and eclectic, and less elitist. Such a general audience has a better response to an artwork that fosters a sense of evolution, either by cognitive means – such as storytelling or interpretation – or sensorial factors – like the build up of audio-visual stimuli – or even a balance of both. Knowing that there will be a culmination/conclusion enhances their interest and focus.

The ordinary exhibition spaces, where visitors can circulate randomly, gather in groups and share experiences, presents a number of challenges to any artwork willing to explore the more intimate and thought provoking aspects of interaction, stimulation and perception. Undivided, intimate attention is virtually impossible to attain, and audible noise can be more severely disruptive than visual disturbance. However such a space can also present some interesting advantages from the entertainment perspective, so a careful balance between both aspects should be considered.

The original goal of both projects, which was loosely translated into the creation of a cinematic selfie that would question the self, was generally attained, based on the informal interviews conducted by the author. However it must be noted that information on the artworks and their intention also played a crucial role in this attainment. Most visitors who ignored the information – available on site – were limited to the entertainment aspect of the artworks. Therefore, such possibilities should also be considered and addressed, in order to provide meaningful experiences.

It is, therefore, safe to conclude that some of the issues and questions behind the motivation for developing these artworks may not have been fully answered, but the process allowed for the realisation that finding deterministic answers was significantly less important than understanding the shared experience. If there should be one final conclusion, it is then the validation that (digital-media) art retains an irreplaceable role in questioning society, promoting critical thinking and stimulating debate, especially when looked upon the conceptual and formal angles, along with its increasingly gained new ground as a new means of socialization and entertainment.

Final Notes


2A self-portrait produced with the front-facing camera of a smartphone or facing a camera operated by the target subject, whose main purpose is to be subsequently shared on social networks.

3 -2480023beb5c



6 -9310669




Carver, C.S. 2012. “Self-awareness” in Handbook of self and identity, edited by Leary, M.R. and Tangney, J.P.: 50-68. The Guilford Press

Casey, N. 2017. “In Praise of Selfies” in Pacific Standard. accessed on the 25th April 2022

Combes, M. 2013. Gilbert Simondon and the Philosophy of the Transindividual. MIT Press.

Deresiewicz, W. 2015. “The neoliberal arts: How college sold its soul to the market” in Harper’s Magazine.

Frey, D. 1978. “Reactions to success and failure in public and private conditions” in Journal of Experimental Social Psychology, 14: 172-179.

Galanter, P. 2014. “Complexism and the Role of Evolutionary Art” in The Art of Artificial Evolution A Handbook on Evolutionary Art and Music, edited by Romero, J. and Machado, P.: 311-333. Springer Berlin.

Gill, R. 2021. Changing the Perfect Picture: Smartphones, Social Media and Appearance Pressures. City University London. accessed 25th April 2022

Groys, B. 2008. Art Power. MIT Press.

Guattari, F. 2000. The Three Ecologies. The Athlone Press.

Hawkins, I and K. Scherr. 2017. “Engaging the CSI effect: The influences of experience-taking, type of evidence, and viewing frequency on juror decision-making” in Journal of Criminal Justice, 49: 45-52.

Hukkelås, H., Mester, R. and Lindseth, F. 2019. “Deepprivacy: A generative adversarial network for face anonymization” in International symposium on visual computing: 565-578. Springer, Cham.

Lipovetsky, G. and J. Serroy. 2016. L’esthétisation du monde: Vivre à l’âge du capitalisme artiste. Folio.

Lopatina, O.L., Y.K. Komleva, Y.V. Gorina, H. Higashida and A.B. Salmina. 2018. “Neurobiological aspects of face recognition: the role of oxytocin” in Frontiers in behavioral neuroscience, 12: 195.

Lotman, Yuri. M. 1990. Universe of the Mind: A Semiotic Theory of Culture. London: IBTauris.

March, E. and T. McBean. 2018. “New evidence shows self-esteem moderates the relationship between narcissism and selfies” in Personality and Individual Differences, 130: 107-111.

Marche, S. 2013. “Sorry, Your Selfie Isn’t Art,” in Esquire, The Culture Blog, July 24, 2103. accessed on 25th April 2022

Markos, E., L.I. Labrecque and G.R. Milne. 2012. “Web 2.0 and Consumers’ Digital Footprint: Managing Privacy and Disclosure Choices in Social Media” in Online Consumer Behavior: 191-216. Routledge.

Marwick, A. and D. Boyd. 2011. “To see and be seen: Celebrity practice on Twitter” in Convergence, 17(2): 139-158.

McCormack, J., O. Bown, A. Dorin, J. McCabe, G. Monro and M. Whitelaw. 2014. “Ten questions concerning generative computer art” in Leonardo 47(2): 135-141.

Moura, J.M. and Ferreira-Lopes, P., 2017. “Generative Face from Random Data, on’How Computers Imagine Humans’” in Proceedings of Artech 2017, the 8th international conference on digital arts: 85-91.

Origgi, G. 2019. Reputation: What it is and why it Matters. Princeton University Press.

Pollock, D. 1995. “Masks and the Semiotics of Identity” inJournal of the Royal Anthropological Institute, 1(3): 581-597.

Rosen, J. 2016. “Examining Preadolescent Television Programming and the Rise of Generation Me” in Elon Journal Of Undergraduate Research In Communications, 7(1).

Simondon, G. 1992. “The Genesis of the Individual” in Incorporations, edited by Crary, J. and Kwinter, S.: 297-319. Zone Books.

Simondon, G. 1964. L’individu et sa genèse physico-biologique. PUF.

Smith, L.F., J.K. Smith and P.P.L. Tinio. 2017. “Time spent viewing art and reading labels” in Psychology of Aesthetics, Creativity, and the Arts, 11(1): 77–85.

Turner, G. 2006. “The Mass Production of Celebrity: ‘Celetoids,’ Reality TV, and the ‘Demotic Turn’” in International Journal of Cultural Studies, 9(2): 153–165.

Twenge, J.M. 2013. “Teaching generation me” in Teaching of Psychology, 40(1): 66-69.

Veiga, P.A. 2017. “Generative theatre of totality” in Journal of Science and Technology of the Arts, 9(3): 33-43.

Veiga P.A., M. Tavares and H. Alvelos. 2017. “How experience, attention and ubiquity economies affect the role of digital media art and artists” in Journal of Spatial and Organizational Dynamics, 5(3):.222-238.

Young, S.M., F. Du, K.K. Dworkis and K.J. Olsen. 2016. “It’s all about all of us: The rise of narcissism and its implications for management control system research” in Journal of Management Accounting Research, 28(1): 39-55

Wang, X., Li, Y., Zhang, H. and Shan, Y. 2021. “Towards real-world blind face restoration with generative facial prior” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition: 9168-9178.