Posthuman and Hyperreal Self on Facebook: AI-Generated Images, Aesthetic Labour, and Everyday Digital Selfhood

Abstract

This article interrogates how an everyday Facebook user in Nigeria adopts AI-generated images as a sustained mode of online and digital self-presentation. This study situates this practice within debates on posthumanity, aesthetic labour, and platform affordance. Drawing on a qualitative case study of a pseudonymised account (“Ada Alika”), this article analyses 25 AI-generated self-images posted between February and September 2025 alongside 25 non-AI-generated images. Using a triangulated methodological approach that combines media archaeology, online observation, and visual thematic analysis, the article describes how generative AI functions as an aesthetic and ontological collaborator in the construction of online selfhood. The results from the analysis are organised around three interconnected themes: “Posthuman and Hyperreal Self”, Aesthetic Labour and AI-mediated Persona”, and High Audience Engagement. The analysis indicates that AI-generated images produce a hyperreal self that exceeds naturalistic representation while remaining socially legible and desirable. These images demand sustained aesthetic labour and align closely with Facebook’s attention economy, resulting in markedly higher engagement than non-AI images. Audience interactions further reveal a level of visual literacy in which distinctions between the real, artificial, and aspirational are fluidly negotiated. By foregrounding a non-Western and non-influencer context, this article extends existing scholarship on AI-mediated selfhood and demonstrates how AI-generated images on Facebook reshapes everyday practices of self-representation and visibility in online digital culture.

1. Introduction

I began this research with a simple curiosity of why AI-generated images were becoming prevalent with an ordinary and everyday Facebook user. I was particularly curious in the contexts that fall outside the commercial and influencer-driven ecosystems that dominate digital media scholarship. As I spent time scrolling through my field site, a Nigerian-based Facebook account I pseudonymised as “Ada Alika”, I found myself drawn into the visuals of the account’s posts. What initially appeared as playful experiments with AI-generated images gradually revealed themselves as part of a larger, ongoing self-making practice. The deeper I scrolled, the more I noticed a shift: Ada Alika’s earlier and naturalistic photos faded into obscurity, both aesthetically and in audience reception, while the later AI-generated portraits radiated with hyperreal luminosity and accumulated the densest clusters of engagement. This encounter guided my methodological and analytical orientation. Rather than treating Ada Alika’s posts as isolated artefacts, I approached her account as a cultural fieldsite, an unfolding narrative where selfhood was negotiated through machinically augmented elements. My position as a researcher was shaped by the embodied practice of scrolling, observing, comparing, and returning, allowing me to witness how her digital self evolved in form and in audience interpretation and platform visibility. The account offered a vivid lens into how ordinary users in non-Western contexts engage with generative AI technologies and aesthetic norms, challenging assumptions that AI-mediated self-fashioning is exclusive to professional content creators.

This study emerges against a backdrop of rapidly expanding scholarship on AI-mediated embodiment, posthumanity, and digital cultures. Researchers have shown that generative AI tools increasingly co-author the body, producing speculative and hyperreal forms that transcend biological limits (Corrêa, 2023; Lee et al., 2023). Others highlight how platforms privilege certain aesthetic conventions, smoothness, symmetry, and saturation, which shape what becomes visible, desirable, and socially validated (Bucher, 2018; Cotter, 2019). Yet when applied more broadly, these discussions remain heavily centred on Western influencer cultures (Leaver et al., 2020), leaving significant gaps in understanding how AI-generated self-images circulate in everyday non-influencer contexts. This article responds to these gaps. Through a triangulated qualitative approach encompassing media archaeology, visual analysis, and platform affordance analysis, I examine how Ada Alika adopts AI-generated images of herself as a form of selfhood.

Specifically, this article attempts to answer the question: How do everyday Facebook users such as Ada Alika employ AI-generated images of herself to construct and negotiate online selfhood? Also, the article also attempts to understand the engagement metrics of AI-generated and non-AI-generated imagery posts on the account. I answered this question through the analysis of 25 AI-generated images of Ada Alika posted on her account between February 2025 and September 2025, and another chronologically selected 25 non-AI-generated posts on the account. It is pertinent to mention that this article did not necessarily compare AI and non-AI images in general. Rather, it explores engagement differences within the same account across two adjacent time periods before and after the adoption of AI-generated imagery.

Consequently, the article presents its results and discussion through three interconnected themes: “Posthuman and Hyperreal Self”, “Aesthetic labour and AI-generated Persona”, and “High Audience Engagement.” Through these themes, I argue that AI-generated images function not merely as beautification conduits but as cultural technologies that reconfigure selfhood, relationality, and the politics of engagement on Facebook. In the sections that follow, I discussed the literature on AI-mediated embodiment, followed by a section on aesthetic labour, visual culture, platform visibility and digital selves. Afterwards, I moved to the methods and procedure before discussing the results, and then the conclusion.

2. AI-Mediated Embodiment and Posthumanity

A rapidly expanding body of scholarship in digital media studies has examined the intimate entanglements between artificial intelligence, digital aesthetics, and contemporary modes of self-representation. Generative AI tools, ranging from facial enhancement applications to full-scale image synthesis, have introduced new forms of visual self-fashioning that radically extend traditional photographic practices. Scholars across posthuman studies, platform studies, and visual culture argue that AI-generated images increasingly produce posthuman embodiments, where systems co-author the body and destabilise the boundaries between the organic and the computational (Gunkel, 2020; Braidotti, 2019; Sushchenko & Yatsenko, 2025; Menon, 2024). These hybrid representations foreground AI as an aesthetic and ontological collaborator that produces bodies that are simultaneously idealised, speculative, and technologically mediated. Literature on generative portraiture reveals that AI images engender hyperreal beauty logics that exceed naturalistic representation, generating highly polished faces, symmetrical bodies, and fantastical visual signatures that would be unattainable through conventional photography (Khoo et al., 2022; Lee et al., 2023; Corrêa, 2023). These aesthetics tend to reproduce a sensibility marked by luminous surfaces, immaculate skin textures, and curvilinear perfection. These visual forms circulate within broader economies of platformised beauty, where digital smoothness, porelessness, and stylised hyper-femininity have become normalised and culturally aspirational (Banet-Weiser, 2018; Hussain et al., 2025; Trammel, 2023). Furthermore, research demonstrates that AI not only extends existing beauty norms but actively generates new aesthetic expectations, reshaping how users imagine and relate to their own bodies (Vargas-Veleda et al., 2025; Corrêa, 2023).

Digital-self such as AI-mediated embodiment is also deeply entangled with broader transformations in digital identity. As scholars (Penny, 2017; Spawforth-Jones, 2024; Tiidenberg & Gómez Cruz, 2015) have argued, digital self-images now function as speculative identity experiments, spaces where individuals rehearse, negotiate, and contest their sense of self through technologically augmented elements. Generative AI intensifies these dynamics, offering users an expanded palette of speculative futures: alternative selves, aspirational and posthuman imaginaries. These machine mediated selves are not merely decorative; they materialise new ontologies of embodiment that shape how individuals understand their identities, social value, and desirability within digital culture.

Despite these significant insights, the existing literature exhibits critical limitations. First, research is overwhelmingly concentrated on Western contexts. This geographic and demographic bias obscures how everyday users, particularly in global majority contexts, engage with AI-generated self-images outside the commercialised economies of influencer culture. Second, while scholarship has documented the emergence of AI beauty filters and portrait apps, far less attention has been paid to how ordinary users incorporate AI-generated images into their long-term selfhood curation practices, especially on older platforms such as Facebook, which remain popular outside Western-centric research contexts. This article responds to these gaps by offering an interpretative grounded analysis of an everyday Facebook user who regularly posts AI-generated portraits as a sustained mode of self-hood construction. By focusing on non-influencer, non-Western, and everyday practices, this study advances current debates on AI-mediated selfhood and expands the empirical scope of research on digital self-presentation.

3. Aesthetic Labour, Visual Culture, Platform Visibility and Digital Selves

A growing body of scholarship highlights how digital self-presentation requires significant aesthetic labour, involving continuous refinement, curation, and creative decision-making (Elias et al., 2017; Duffy & Hund, 2019; Duffy et al., 2021). This labour extends beyond influencers to everyday social media users who must navigate shifting norms of visual culture, and digitally mediated beauty standards (Abidin, 2016; Graham et al., 2017). AI-generated images intensify these dynamics by adding new layers of creative and technical decision-making: selecting prompts, editing outputs, manipulating stylistic elements, and integrating AI-produced images into a coherent online persona. Despite these emerging contributions, existing literature has focused largely on influencer cultures, professionalised content creation, or commercial beauty practices (K. Harris & Willis, 2025; Ghaffari, 2023). Far less is known about how ordinary users engage in aesthetic labour through AI-generated images, especially when such labour blends fantasy, glamour, and self-reinvention. Additionally, scholarship rarely explores the intersection of generative AI with everyday visual cultures on platforms like Facebook, which remain understudied compared to Instagram, TikTok, and Snapchat (Hernández-Serrano et al., 2022; E. Harris & Bardey, 2019). By comparing and analysing the AI-generated images and non-AI-generated images of a non-professional influencer on Facebook, this article addresses these gaps and provides insight into how ordinary individuals adopt, negotiate, and sustain visual digital selfhood.

Platform studies scholarship has emphasised that identity performances online are shaped not only by user intention but also by platform logics and infrastructures that structure visibility, engagement, and circulation (Nieborg & Poell, 2023; Bucher, 2018; Van Dijck et al., 2018). Platforms prioritise content that is visually striking and emotionally charged, thereby creating an environment in which certain forms of self-presentation, especially polished, stylised, or hyper-aestheticized images, receive amplified visibility (Bishop, 2025; Cotter, 2019). This dynamic is further intensified by the platform’s dependence on metrics such as likes, reactions, comments, and ranking systems, which collectively shape how users understand and adjust their self-presentation strategies online (Theocharis et al., 2023). Although scholars have begun exploring the relationship between AI-generated imagery and platform economies (Vrabič Dežman, 2024; Wang & Liang, 2023; Hausken, 2024), little is known about how AI generated self-images circulate in non-influencer and non-Western contexts. Moreover, while prior research highlights how platforms reward visual spectacle, there is limited empirical work on how AI-generated images oscillate between realism, fantasy, and self-idealisation. This article contributes to filling this gap by examining how Facebook’s platform logics shape audience engagement with AI-generated self-images and how Facebook users engage and interact with these images.

4. Methodology

This study adopts a triangulated qualitative approach that integrates media archaeology (Parikka, 2013; Malloy, 2016), online observational approach and visual thematic analysis (Heng, 2019) to examine how AI-generated self-images are produced, circulated, and interpreted within a Facebook account I pseudonymized this Facebook account as “Ada Alika.” In this article, I employ qualitative triangulation as a strategy for analytic accountability across distinct forms of data. Media archaeology informed the longitudinal reading of the account, enabling the identification of a temporal shift from naturalistic to AI-generated self-images and situating this transition within broader histories of digital beautification and platformed selfhood. With visual thematic analysis, I provide systematic evidence for recurring aesthetic features (e.g., glamour, luminosity, and fantasy) and supported claims about posthuman and hyperreal self-representation. While the platform-based online observation enabled me to generate evidence from captions, comments, and interactional tone, which grounds the commenters’ interpretations, recognition, and validation. With these approaches, I anchored theoretical interpretation in observable visual, temporal, interactional, and metric-based data. The table below (

Table 1. Methods, Data, and Analytic Contributions.

Methodological Approaches	Data Extracted	Analytical Contribution
Media archaeology	Timeline of posts; shift in image type	Identified transition to AI imagery and its durability over time
Visual thematic analysis	Recurrent visual features and motifs	Supported claims about hyperreal and posthuman aesthetics
Online digital Observation	Captions, comments, affective tone	Grounded claims about audience interpretation and validation

) summarises the methodological contributions.

Ada Alika is a Nigerian lady in her mid-30s, based in Lagos, Nigeria, an urban, commercially dense media environment. In 2025, 25.85 per cent of Nigerians use Facebook, and the portion of the population using the social network is projected to increase to 29 per cent in 2026 (Statista, 2025b). It is pertinent to mention that Facebook has one of its African Offices in Lagos (Meta, 2020). The platform is widely used across adult age cohorts, especially those aged 25–34 (Statista, 2025a), making it a key site for everyday sociality and public-facing self-presentation. Importantly, Facebook in Nigeria is not only a space for interpersonal connection but also a venue for small businesses, microentrepreneurs, and informal commerce (Udenze & Aduba, 2020; Nyekwere et al., 2013). In other words, Facebook functions as a mixed social-commercial ecology in which visibility, reactions, and shares carry social as well as material value.

Because the account is owned by someone whom I have met and I know, I can confirm that the generated AI images I studied are images of herself. In addition, the owner of the account identifies as she/her. As a consequence, I will refer to Ada Alika as She/Her. More broadly, the focus of the research is on understanding how the account constructs digitally mediated identities and how platform infrastructures shape the aesthetic and affective logics of self-representation. More broadly, it also focuses on audience engagement with AI-generated self-images. Consequently, I treated the account as a cultural fieldsite where selfhood is performed and negotiated. In conducting this study, I followed an iterative and systematic data collection process designed to capture both the visual and contextual dimensions of the AI-generated self-images posted on the public Facebook account.

This study did not solicit from Ada Alika detailed information regarding the technical processes or workflows underlying the AI-generated images, such as whether the images were produced through text-to-image systems, the specific tools employed, or the use of enhancement filters. Rather my identification of Ada Alika’s AI-generated images was based on my media-literate comparative analysis of earlier non-synthetic images and later posts. I used a set of observable visual indicators. These indicators included: (a) repeated hyper-smooth skin textures lacking photographic grain; (b) indeterminate or implausible background spaces without stable spatial referents; (c) proportion distortions or exaggerated facial symmetry inconsistent with earlier images; and (d) localised artifacting around hair, jewellery, necklines, or edges of the body. I classify images as AI-generated only when multiple indicators co-occurred. However, I acknowledge that this inferential approach may result in false positives, particularly given advances in image editing and enhancement tools.

My first step was to create a corpus of 25 publicly accessible AI-generated imagery posts published on the account between February 2025 and September 2025. The posts were included based solely on the presence of AI-generated self-imagery, irrespective of audience engagement levels. To contextualise engagement patterns, I did a comparison set of 25 public accessible non-AI portrait images posted immediately prior to the adoption of AI-generated imagery. These posts were selected chronologically and independent of engagement metrics. This approach ensures transparency in data provenance and minimises intrusion into the user’s private online activity. I made no attempts to bypass privacy settings, access restricted content, or engage in covert observation.

For the 25 AI-generated images in the Facebook post, I coded the material manually through an iterative and interpretive framework grounded in scholarship on visual culture. Employing approaches akin to Heng’s (2019) visual thematic analysis techniques, I analysed and coded each AI-generated image for symbolic content and representational elements. Rather than saving or reproducing the images directly, I produced an analytic fieldnotes that captured the key visual and aesthetic qualities of each of the AI-generated images. My analysis followed a phased qualitative coding process that combined inductive pattern identification with theoretically informed interpretation. In the first phase, I conducted open coding across the AI-generated images, attending to recurrent visual features (e.g., facial symmetry, skin texture, lighting, pose), caption language, and dominant interactional patterns in comments. This process generated a set of descriptive codes, including “smooth skin”, “fantastical background”, “regal pose”, “aspirational caption”, “awe”, “playful disbelief’, and “affirmation”. These codes were applied iteratively across posts to ensure analytic consistency and to identify recurring aesthetic and interactional patterns. In the second phase, I grouped these descriptive codes into higher-order categories through constant comparison. For example, codes such as “smooth skin”, “glossy finish”, and “symmetrical facial features” clustered under the broader category beauty, while “otherworldly”, “lighting”, “unreal environments”, and “non-naturalistic proportions” informed the category fantasy. These clustered categories provided the empirical foundation for the thematic articulation of the posthuman and hyperreal self. It captures how AI-generated imagery produced figures that exceed human realism while remaining socially legible. In the third phase, I refined and consolidated themes in dialogue with the manuscript’s theoretical framework. Codes associated with captions emphasising self-worth, transformation, or affirmation (e.g., “becoming,” “soft life,” “my era”), alongside supportive and affirming audience responses, were grouped under self-celebration and empowerment. These patterns informed the theme aesthetic labour and the AI-generated persona, foregrounding the work involved in curating, sustaining, and iteratively refining a coherent posthuman aesthetic across posts. In parallel, codes emphasising luxury aesthetics, stylized elegance, and performative excess were consolidated under glamour, highlighting the laborious and aspirational dimensions of AI-mediated self-presentation.

Engagement constituted another layer of coding and analysis. I collected the engagement data on the AI and non-AI image posts using Facebook visible metrics available. For each post, I recorded the total number of reactions, comments and shares. I observed how commenters interacted with the AI-generated images, whether through compliments, humour, admiration, critique, or expressions of desire. I noted instances where commenters appeared to interpret the images as realistic portraits, artistic renderings, playful experiments, or aspirational versions of self. This contextual layer helped me situate the images within the affective and social dynamics of the platform. The comment threads were analytically central to the theme of high audience engagement. Given the qualitative and single-account design of this study, I reported engagement metrics descriptively rather than inferential statistical testing. The purpose of this comparison is not to establish causal effects, but to render visible relative differences in interactional intensity across content types within the same platform ecology.

Further, I coded emotional and affective tone using categorical variables derived from both visual cues and audience responses. These included values such as “admiration”, “awe”, “playfulness”, “aspirational desire”, and “scepticism”. I inferred emotional tone not just from subjective impression alone, but from observable markers, including emoji use, lexical choices in comments, repetition of evaluative phrases (e.g., “unreal,” “queen,” “this is too much”). These affective patterns were also analytically central to the theme of high audience engagement, demonstrating how AI-generated personas elicited sustained and participatory interaction.

Further, for each AI-generated image, I also wrote a set of interpretive reflections that captured my initial analytical impressions. These reflections addressed the selfhood narrative that the image seemed to construct. This reflexive engagement allowed my analysis to evolve alongside the data. Together, this data collection and analysis process allowed me to approach the account systematically and contextually. It enabled a nuanced interpretation of how AI-generated images mediate self-representation on Facebook and how audiences engage with them. In essence, these coding phases ensured that the three core themes, “Posthuman and Hyperreal Self”, “Aesthetic Labour and AI-generated Persona”, and “High Audience Engagement”, are constructed from systematic engagement with the data, while remaining attentive to the relational, interpretive, and platform-mediated nature of visual meaning-making on social media. The methodological choices in this study reflect a broader theoretical engagement with scholarship on digital identity and AI-mediated selfhood. By triangulating these methodological approaches, I aim to situate the user’s AI-generated self-images within ongoing debates about posthumanity and the politics of self-presentation in the online sphere. This approach ensures that the article moves beyond description to offer a theoretically informed interpretation of how AI-generated images shape contemporary practices of self-representation on Facebook.

I adhered strictly to contemporary ethical standards for research involving social media data. Because the content analysed is public, formal consent may not necessarily be required. However, I received her consent to study the account. In line with online research guidelines such as the British Psychological Society (2021), the account owner’s real name is not used. Instead, the user is anonymised through a pseudonym, and no direct quotations or verbatim comments appear in the final publication. All descriptions of images and comments were paraphrased to prevent reverse identification and minimise reputational or whatsoever harm to the individual. No images are reproduced in the article; instead, findings are presented through thick thematic descriptions. In addition, I did not attempt to infer psychological states, vulnerabilities, or private information from the images or interactions. In the next section, I present and discuss the themes and results from the analysis.

5. Results and Discussion

5.1. Posthuman and Hyperreal Self

Across the 25 AI-generated images, I observed a dominant and persistent pattern in which the user curated a hyperreal, aesthetically perfected version of herself, a phenomenon that aligns closely with what scholars (Braidotti, 2019; Gunkel, 2020) argue about posthuman embodiment. As I analysed the images, I discovered that the AI-generated portraits repeatedly exceeded naturalistic representation. The rendered skin appeared poreless, luminous, and immaculately smooth; the facial planes were subtly softened, creating a symmetry that did not exist to the same extent in the user’s unfiltered images. In addition, the lighting effects appeared engineered to maximise a sense of ethereality, while digital makeup was precise, saturated and impossibly seamless. It enhanced the face beyond what one could feasibly achieve through conventional cosmetic practices. These visual modifications were not incidental. They cohered into an overarching logic that matched AI aesthetics, a platform-rewarded visual style marked by hyperreal, polished, and highly curated representations. As I compared the images across posts, I saw this aesthetic functioning as a unifying principle. It guided the representational choices Ada Alika made and shaped the technical adjustments that the AI system automatically layered onto the body. For instance, a portrait depicts Ada Alika’s face rendered with exaggerated smoothness and symmetrical precision. The skin surface appears poreless and evenly lit, with no visible texture or shadow. The eyes are unusually large and luminous, reflecting an artificial light source not present in the background. The background itself is indeterminate, neither indoor nor outdoor, which suggests no fixed spatial referent. Ada Alika describes the portraits as living a soft life. Commenters repeatedly describe the image as unreal, not human, and AI vibes, yet respond with admiration rather than scepticism. This convergence of visual excess and audience recognition grounds the interpretation of a hyperreal persona: a self-presentation that exceeds human realism while remaining socially legible and desirable.

Importantly, this aesthetic did not operate at the level of surface decoration alone. Instead, it actively recomposed the user’s self-presentation. The images revealed a gradual distancing from biological corporeality in favour of what could be called a computationally augmented self, a direction that reflects researchers (Lee et al., 2023; Corrêa, 2023) position that posthuman embodiment emerges when AI systems co-produce the conditions of bodily representation. As I engaged with the dataset, I recognised that Ada Alika was not simply enhancing existing features; she was experimenting with an alternative corporeal template. The resulting persona was not a fictional character but a speculative reconfiguration of the self: an aspirational, machinically informed body that remained recognisably hers while also transcending her biological limits. The consistency of this hyperreal persona across posts suggested that the user was not engaging in one-off play but participating in a sustained selfhood project. However, there is an image of Ada Alika that presents a stylized portrait with subtle inconsistencies: her earrings do not align symmetrically, hair strands blend unnaturally into the background, and a faint distortion is visible around the neckline. In the image description, Ada Alika asserts that it is still her. Several commenters explicitly note the artificiality (This AI is learning fast), while others playfully affirm identity (It is still you though). The lack of correction or clarification by the account owner suggests a simulated orientation, where authenticity is neither claimed nor denied but collectively negotiated through interaction.

The recurrence of similar stylistic decisions, symmetry, luminosity, flawless texture, and subtle posthuman markers such as digitally elongated necklines or smoothed bone structures. This indicates an ongoing negotiation of what she wanted her digital self to become. As I examined this repetition, I saw a pattern akin to Tiidenberg and Gómez Cruz’s (2015) argument about selfhood. However, in the present state it is co-constructed; a situation in which human intentionality entangles with platform logics. Furthermore, the results echoed the literature suggesting that AI functions as more than a representational tool. Instead, it operates as an ontological collaborator, one that materially shapes how individuals envision, interpret, and circulate their identities online (Spawforth-Jones, 2024). These results affirm and extend the argument I developed in the literature review: everyday users, far beyond the realm of professional influencers, are increasingly engaging in what can be understood as posthuman aesthetic labour. This labour involves the continuous negotiation of selfhood through AI-mediated images that blur the boundaries between the natural and the artificial, the biological and the computational, the self and its machinic extension.

As I reflected on the dataset, I recognised that this labour also introduced a feedback loop: the more the user engaged with AI-generated self-images, the more her images appeared to internalise the hyperreal aesthetic as a desirable standard for self-presentation. Her captions or descriptions of the images suggested an increasing comfort with, and even celebration of, this mechanically enhanced version of herself. For instance, there were recurring comments that portrayed audience recognition. These paraphrased comments also describe Ada Alika’s posthumanness. I presented some of the comments that were rendered colloquially and in Nigerian Pidgin English: “This babe looks like the one from last week o, just another level. Exactly. It’s like she’s upgrading every time.” “Same person, different version”. I like this thing.” This exchange reflects audience recognition of serial continuity and incremental transformation. The use of terms such as babe, another level, and upgrading situates world-building within everyday Nigerian digital vernacular, emphasising progression rather than rupture. Further, there was this AI-generated image that depicted temporal and spatial coherence. Ada Alika describes the image as another version of herself in another day with the same energy. Some of the top comments from this post are: “This page don turn like its own world, same vibe, new mood every time. You dey carry us enter this your aesthetic world small small. Once you see one post, you already know it’s from this page.” Here, world-building is articulated through temporal layering (another day) and spatial abstraction (another version of…). It is also articulated through familiar Nigerian expressions (the same energy, don turn), with audiences explicitly articulating the sense of a coherent, self-contained world. The interaction shows how platform users participate in sustaining the imagined environment through interpretive alignment. Taken together, the results reveal a complex dynamic in which AI systems participate in the production of hyperreal selves that are not merely enhanced but ontologically reconfigured. This result demonstrates that Ada Alika’s engagement with AI portraiture constitutes an active form of posthuman world-building, one in which selfhood is not fixed but continually re-authored through collaborative interactions between human and platform interactions.

5.2. Aesthetic Labour and AI-Generated Persona

As I moved deeper into the dataset, I observed that the user engaged in sustained and intentional aesthetic labour, a pattern that strongly supports Duffy and Hund’s (2019) claim that digital self-presentation requires ongoing curatorial work, even for individuals who do not participate in professional influencer cultures, such as in the present case. In the context of this result, I operationalised aesthetic labour as labour that is observable at the level of platform practice rather than image production. I refer to such labour as sustained curatorial work, which includes the selection and sequencing of images, consistency of visual style across posts, captioning strategies, and the management of audience interaction. It does not include unobservable activities such as prompt design, iterative generation, or post-processing. Ada Alika’s posts displayed a striking degree of aesthetic consistency, suggesting that she was not merely experimenting with AI-enhanced imagery but actively cultivating a stable digital persona. This became evident as I coded the visual attributes across images and noticed recurring stylistic, chromatic, and symbolic motifs that collectively established a coherent personal brand. One of the first elements that stood out to me was the user’s consistent use of colour. Across the dataset, she selected jewel tones, pink glamour hues, and other saturated palettes that created a mood of fantasy, opulence, and emotional warmth. These colours did more than simply frame the user attractively; they contributed to what I interpret as a deliberate crafting of ambiance, an affective environment in which the hyperreal self could appear credible and desirable. The chromatic consistency functioned as a visual anchor, a way of guiding viewer expectations and creating a recognisable style that persisted regardless of changes in backdrop or pose.

In addition to colour, I noticed a pattern of recurrent symbolic environments. Ada Alika frequently appeared within ethereal, celestial, or regal spaces, settings that AI tools often generate when prompted with glamour or fantasy aesthetics. These backdrops, filled with soft glows, shimmering textures, and ornate ornamentation, positioned her within a quasi-mythic visual universe. As I analysed these environments, I saw them operating as interpretive cues: they signalled that the persona being curated was not meant to be read as strictly realistic but as aspirational and larger-than-life. She repeatedly returned to these symbolic spaces, demonstrating aesthetic preference and a deeper orientation toward self-construction through fantasy-infused visual worlds. The labour involved in producing these posts also manifested in her stylised poses and facial expressions. As I compared images, I noticed that she repeatedly adopted poses that maximised symmetry, elegance, and emotional resonance. These poses were often complemented by smooth textures and subtly symmetrical facial features. For instance, one of Ada Alika’s posts features a recurring aesthetic seen across multiple AI images: neutral pastel backgrounds, minimalist jewellery, and a consistent facial structure that persists across different outfits. Comment threads link the image to earlier posts (This looks like the last one but upgraded). The repetition of visual motifs across posts constructs a recognisable, internally coherent aesthetic universe. This supports the claim of world-building, not as narrative continuity, but as visual consistency that allows audiences to recognise and inhabit an evolving aesthetic environment.

As I reflected on the dataset, I inferred that Ada Alika’s curatorial decisions were driven by affective, aspirational, and self-focused motivations. Her engagement with AI portraiture was part of an ongoing personal project, one in which digital self-making became a space for experimentation, play, and self-elevation. Crucially, the user’s aesthetic decisions also revealed a tacit attunement to what the Facebook logic tends to attract when compared to her natural images, as we would see in the next theme. Further, her reliance on highly stylized, vibrant, and emotionally expressive imagery suggests an embodied knowledge of how to speak in the platform’s preferred visual language. This orientation is consistent with research showing that users often internalise platform norms through repeated interaction, eventually developing a sense of what works without necessarily articulating it in technical terms.

These findings extend and complicate existing scholarship in several ways. First, they demonstrate that aesthetic labour is not confined to influencers, beauty content creators, or highly curated Instagram subcultures. Rather, ordinary social media users also engage in forms of digital authorship that require time, attention, and creative investment. As I considered this point, I recognised that the dataset offered a valuable counterexample to the influencer-centric focus that dominates much Western scholarship. By documenting the practices of an everyday user, this article highlights the broader cultural diffusion of aesthetic labour as a mode of selfhood construction.

5.3. High Audience Engagement

As I analysed the circulation patterns of the AI-generated portraits, I observed that Facebook’s logic shaped audience engagement in clear and measurable ways. With hyperreal and glamorous features, AI images received higher and visible engagement within Ada Alika’s account’s network and this pattern is consistent with attention-economy incentives. In coding the engagement metrics, I found that audience responses to the AI-generated images tended to cluster into three dominant forms: admiration, playful disbelief, and interpretive validation. Many commenters expressed direct praise (You look stunning; Absolutely gorgeous), using language that reinforced the glamour aesthetic embedded in the images. Others responded through humorous or teasing disbelief (Is this really you? AI did magic here), indicating that they recognised the enhanced, hyperreal qualities of the portraits. A third group of commenters offered interpretive affirmations, with remarks such as, This is art, which reframed the images not simply as representations of the self but as creative artefacts worthy of aesthetic appreciation. As I reviewed these comments, I became increasingly aware that the audience was not treating the images as straightforward depictions of reality. Instead, they engaged them as hybrid objects, both aspirational and authentic, artificial yet meaningful. This interpretive flexibility aligns with Hernández-Serrano et al.’s (2022) findings that digital audiences increasingly read images through layered modes of realism, affect, and aesthetic valuation. Ada Alika’s AI-generated images seemed to invite this layered reading, encouraging viewers to oscillate between appreciating the technological artistry and affirming her selfhood through these machinically mediated forms.

In addition, I noticed a subtle but important shift: rather than policing the boundary between real and artificial, the commenters appeared comfortable suspending that distinction. They responded to the portraits as if they were simultaneously representations of the user and aesthetic creations with their own integrity. This interpretive suspension suggests that Ada Alika’s audience approached digital images through a hyperreal lens, in which the value of an image lies not solely in its fidelity to the physical body but in its capacity to evoke affective and aesthetic resonance. Besides these patterns, the AI-generated images consistently garnered a considerable amount of reactions.

Although modest in absolute terms, within the context of the account’s typical engagement history, these engagements represent a marked elevation in visibility. The presence of shares is especially significant, as sharing constitutes a higher-order form of interaction that signals not only appreciation but also a willingness to circulate the image beyond its immediate social boundaries. This amplification effect suggests that AI-generated images attract more attention and mobilise the audience to participate in expanding Ada Alika’s reach and visibility.

When I compared her AI-generated images to her natural photographs posted prior to her adoption of AI-generated portraiture. I observed a striking contrast in patterns of audience engagement (See

Table 2. Descriptive engagement metrics for AI-generated and non-AI posts on Ada Alika’s account.

AI-generated Image (N = 25)	Reactions	320–980
	Comments	55–312
	Shares	24–146
Non-AI-generated images (N = 25)	Reactions	97–258
	Comments	15–82
	Shares	6–38

). These earlier, naturalistic images, which depicted Ada Alika in everyday settings and without digital enhancement, accumulated far fewer engagements. Several of them registered minimal engagements. The minimal engagement, especially when set against the rich engagements of her later AI-generated images, was revealing. It suggested that the audience either overlooked or deprioritized these natural images within the visual economy of Facebook. In reflecting on this contrast, I began to see how the platform’s cultural logics converge to privilege certain forms of visual production over others. The small engagements surrounding the older natural photographs indicates that natural images, despite their indexical connection to the user’s corporeal self, do not command the same attention as the aesthetically intensified portraits produced through AI tools. While differences in engagement are visible across AI and non-AI posts, this article attributes these patterns to the platform logic. Facebook’s content-ranking systems are opaque and shaped by multiple interacting factors, including posting frequency, network structure, and prior audience behaviour. Rather than claiming algorithmic causality, I interpret the observed engagement patterns as platform-visible outcomes, signals that reflect how AI-mediated imagery aligns with existing attention economies and affective cultures on Facebook.

Further, the elevated engagement the AI image received indicates that AI-generated content possesses the capacity to attract attention and foster connection. By examining these patterns, I show that visibility on Facebook is not merely a function of personal networks or posting frequency but a product of how well the aesthetic and affective qualities of content align with the platform’s logic priorities and audience expectations. However, unlike other AI-generated images, an AI-generated image of Ada Alika has an over-stylized background resembling fantasy artwork and exaggerated facial proportions that deviate from earlier motifs. The comments on the image are sparse and largely limited to emojis. The absence of sustained interaction suggests that not all AI-generated imagery produces heightened engagement, which challenges the claim about AI aesthetics and reinforces the importance of stylistic alignment with audience expectations.

Nonetheless, I acknowledge that because the non-AI posts are drawn from the period immediately preceding the adoption of AI imagery, observed engagement differences may also reflect temporal factors such as audience growth, changes in posting cadence, seasonal effects, or broader network dynamics rather than content type alone. As such, this result should be interpreted as descriptive of within-account engagement patterns across adjacent periods, not necessarily as causal evidence of AI imagery effects.

6. Conclusions

This study has examined how an everyday Facebook user, pseudonymised as Ada Alika, adopts AI-generated images as a sustained mode of digital self-presentation. It highlights the complex entanglements among posthumanity, aesthetics, and platform logics of selfhood-making. Compared to her natural images, data indicate that Ada Alika’s AI-generated images were not merely enhanced images but posthuman aesthetic formulations in which AI functioned as an ontological and creative collaborator. Her use of AI-generated imagery produced a speculative, hyperreal persona that operated beyond biological corporeality, demonstrating how everyday users now engage in selfhood projects that extend into mechanically augmented futures. Ada Alika’s practices of producing and engaging with AI-generated images of herself constituted meaningful aesthetic labour, challenging scholarship that positions such labour as the domain of influencers or professionalised content creators. Instead, this case demonstrates how ordinary users participate in AI-shaped selfhood work that blends fantasy, aspiration, and emotional expression. Audience engagement further demonstrated the relational dimensions of these practices. Reactions and comments affirmed the hyperreal persona and also embraced its interpretive ambiguity. This reflects a hyperreal mode of visual literacy in which boundaries between the real, the aesthetic, and the technological are fluidly negotiated. The stark contrast between the visibility of AI-generated portraits and naturalistic images underscores how platform logic rewards engagements and shapes both the production and reception of digital selves.

Further, these results complicate traditional assumptions about authenticity in social media environments. It suggests that AI-generated images do not necessarily undermine the user’s perceived self; instead, they can enhance it by enabling the user to inhabit a more aspirational, stylized, and amplified version of herself. The audience, in turn, affirms this expanded self, participating in what can be understood as a collaborative legitimization of the AI-mediated self. In essence, these results demonstrate that the circulation and reception of AI-generated images depend on a dynamic interplay between Ada Alika’s intention and the audience’s interpretation. Notwithstanding these contributions, this article acknowledges some limitations. The study did not have access to direct information about the technical processes underlying the AI-generated images, including the specific tools, workflows, or enhancement methods used. As a result, inferences about the AI-generated image were based on the researcher’s media-literate comparative analysis of earlier and more recent images, which may introduce interpretive subjectivity. Further, while the exploration of single-account design does not allow causal claims about algorithmic preference, the comparative engagement patterns highlight how AI-generated images align with platform-visible forms of attention.

Taken together, this research expands existing literature by foregrounding a non-Western and non-influencer contexts in which AI mediates selfhood construction. It demonstrates that AI-generated portraits have become culturally meaningful artefacts through which individuals negotiate selfhood, visibility, and social value. More broadly, the study highlights how generative AI reshapes the visuality of digital selfhood and the interpretive, and relational practices through which these identities circulate. As generative AI becomes increasingly embedded in everyday media practices, researchers must continue to interrogate how such technologies reconfigure embodiment, aspiration, and the politics of representation across diverse sociocultural environments.