As large language models (LLMs) with chat interfaces such as ChatGPT have become available to the public, with ChatGPT reportedly amassing more than 400 million active users worldwide, social media has become host to a proliferation of posts and dedicated communities featuring lay people’s experiments with prompt engineering. While much has been written about prompting as a specialized skill, there is currently very little research on the socio-cultural practices of prompting and the online communities dedicated to these activities and little is known about how non-experts are using prompts to shape their interactions with LLMs. With this conference paper I aim to contribute to the understanding of LLM prompting as a vernacular practice in digital culture. Combining the method of digital ethnography of the proliferation and modification of a prompt known as "the eigenprompt" on the social media platform X.com, this paper explores how the how assumptions and ideals about LLM intelligence, interiority, and emotion are expressed through the practice of prompting, and how ideals and preferences about LLMs are negotiated and iteratively reformulated throughout online communities. Finally, I argue that the eigenprompt and similar vernacular prompts should be seen as examples of algorithmic folklore that contribute to the construction of new ways of being and relating to digital others.

As platforms like Instagram shift from chronological feeds to algorithm-driven models, visibility is no longer an organic byproduct of social interactions but an ontological condition structured by opaque computational processes. Existing literature on digital influence—centered on authenticity, visibility management, and commercialization—fails to account for how platforms actively shape what is recognized as influential, valuable, or authentic. We demonstrate how algorithmic infrastructures rupture traditional influencer studies, exposing the limits of frameworks that conceptualize influence as human-centered. Instead, we propose that influencing the digital requires acting for, on behalf of, and through the digital itself. We introduce three interrelated modes to capture this transformation: computational authenticity, algorithmic entrepreneurship, and digital visibility. These categories illustrate how influencers must strategically engage with recommendation systems, optimize engagement metrics, and navigate platform affordances to maintain relevance. This study challenges human-centric models and highlights the structural conditions governing visibility by repositioning digital influence as an ontological entanglement between influencers, platforms, and algorithmic infrastructures. We argue that digital influence is not simply mediated by technology but co-produced by sociotechnical arrangements that define who and what gains recognition in algorithmic environments. This shift demands new conceptual tools to understand how influence is actively negotiated within digital infrastructures rather than merely performed for human audiences.

This study examines how algorithmic control intersects with gendered experiences in China’s food delivery industry, focusing on female riders’ perceptions of algorithms and their marginalization within the platformized gig economy. Through participant observation and semi-structured interviews with riders, we reveal that female riders often internalize algorithmic inequalities as personal limitations, attributing disparities in work outcomes to their gender identity rather than systemic biases. Female riders interpret algorithms as neutral and rational, despite the structural biases embedded in platform designs. Additionally, we find that this perception is reinforced by their limited access to informational networks dominated by male riders, who dominate shared physical and digital spaces. A pervasive misogynistic workplace culture, characterized by the degradation of women’s work and sexual harassment, further excludes female riders from critical information and resistance strategies. Our findings reveal how gender-unequal sociocultural contexts, precarious labor conditions, and opaque algorithmic infrastructures intersect to reproduce structural inequality. By highlighting the gendered dimensions of algorithmic control, this study contributes to broader understandings of how technology and sociocultural dynamics shape labor practices and perpetuate marginalization in the gig economy.

In an increasingly algorithm-driven digital landscape, personalization systems are often praised as almost magical entities capable of "knowing us better than we know ourselves". This study takes TikTok as a privileged object of investigation to explore how users engage with these algorithms that seek to "know" them. Adopting a qualitative, empirical approach, we conducted 20 semi-structured interviews with TikTok users in Brazil to explore their everyday experiences, perceptions, and imaginations around algorithmic recommendations. We argue that TikTok’s algorithm-centered infrastructure shapes a unique user experience, where navigating the platform is essentially about engaging in a relationship with the algorithm itself. Our participants not only demonstrated a high algorithmic awareness but developed informal theories and strategies to “train” their systems in an effort to refine recommendation accuracy. As these users experienced increasingly personalized content — deeply attuned to their emotions and personal experiences — some expressed a sense of algorithmic intimacy. Through their clicks, likes, searches, and interactions, they felt as if “sharing a secret with the algorithm”, revealing personal, sensitive, and sometimes even embarrassing preferences. In this preliminary analysis, we discuss how TikTok’s “For You” model fosters a new form of intimacy mediated by algorithms, data collection, and commercial interests. While the transformation of intimacy is not a new topic in internet studies, this paper shifts the focus from self-exposure on social media to the cultivation of an intimate relationship between users and algorithms, offering new insights into the affective and subjective dimensions of algorithmic personalization.