Co2gnition is a master’s thesis and an autoethnographic experiment on how short-form video shapes thought and agency. Over 40 days it turns everyday scrolling into a method, a diary of Instagram Reels, and introduces the concept of co-cognition — a state in which human cognition becomes entangled with algorithmic and AI systems.

Full thesis text available on request by e-mail

PREFACE. I LOVE INTERNET, IT’S FUCKING DISGUSTING

When I first conceived this work, I anticipated that its connection to my artistic practice would center on my process of transforming visual data—images and footage sourced online—into short video novels through artificial intelligence. However, the intersection revealed itself elsewhere: in an image of rain.

A few months ago, I stumbled upon Jia Tolentino’s Trick Mirror. Her articulation of her Internet experience resonated deeply with me. Tolentino recounts how her online presence evolved alongside her, from writing five hundred words a day on LiveJournal at twelve, to uploading photos on MySpace at fifteen. By her mid-twenties, the Internet had become an essential part of her livelihood. She describes her adult reality as inseparable from what she terms a “feverish, electric, unlivable hell” of relentless and obligatory connection (Tolentino, 2019, 10). At one point, she writes: “I’ll sit there like a rat pressing the lever, like a woman repeatedly hitting myself on the forehead with a hammer, masturbating through the nightmare until I finally catch the gasoline whiff of a good meme” (Ibid.).

I have rarely encountered a more precise depiction of the scrolling experience in writing. Yet, while this description feels strikingly accurate, it reveals little that is new, especially for my generation. Trick Mirror was published in 2019, before terms like brain rot or AI-slop entered common cultural shorthand. The physical, almost masochistic metaphors Tolentino employs seem to anticipate the intensification of algorithmic recommendation systems and AI-generated content. Reading the reflections of someone whose formative years coincided with early user-driven Internet platforms such as GeoCities, I felt a strange mix of envy and nostalgia for a version of the Internet I never experienced. Moreover, I felt fear. If, in 2019, Tolentino described her experience as a search for “the gasoline whiff of a good meme,” then what, I began to ask myself, is happening to me now?

I have been carrying this question with me for several days. The answer came unexpectedly after one of my routine sessions of scrolling through Instagram Reels. I record it here verbatim, as I originally wrote it in my iPhone notes:

I remember feeling ashamed after writing this. I even consulted ChatGPT to determine whether it was ethical to express myself in this manner. Yet, this fragment unexpectedly came to encapsulate much of what shaped the entire process of this thesis: my simultaneous exhaustion, disgust, and fascination with the current state of the internet; the apocalyptic sensibility of a person from Russia living in emigration, isolated from close friends against the background of an unfolding geopolitical catastrophe; my love for technology and interest in artificial intelligence; my artistic practice centered on generative visual content; fourteen years on Instagram, three different accounts, and a period of early influencer activity before the age of fifteen. It also includes the experience—explicitly mentioned in the text—of forming an artistic identity while studying at an art school in an exceptionally privileged context: an art school in Switzerland, once again against the backdrop of a global crisis.

At the time of writing this, I did not imagine that this text would find a place in my thesis. However, while revisiting my notes in an attempt to crystallise how I arrived at this project—this mental state, this cloud of experiences and tags—I rediscovered this entry. I realized that its key images were rain, flood, dumpsters, and cockroaches. I wrote this text in October. In early December, independently, I arrived at the narrative arc for my artistic project, which I was developing in parallel.

In this work, two AI-generated human-like entities, each confined to a separate large-scale vertical screen, simulated waiting at a railway station. At a certain point, it begins to rain. The male entity repeatedly asks the female one: “Is it going to be over?” Each time, he receives a different form of negation. After a while, he tries again: “Sorry, I’m still wondering if it is over?” At that moment, the female entity turns toward him, and powerful streams of water collapse onto his body and submerge him. While he remains underwater, she monotonously repeats:

It’s not going to be over. Everything has never ended at once. Nothing really “ends” cleanly. Nothing needs to end. There is no set time when things are over.

Only when the male entity, choking, begins to repeat these phrases after her does the water recede.

When I developed this narrative, I had completely forgotten about the diary entry quoted above. Only now have I noticed this resonance. What is perhaps even more telling is that these phrases originate from an experimental dialogue with ChatGPT, in which I decided to play a deliberately absurd game—“Are we there yet?”—a game I used to play during long car rides with my parents. After repeating the question multiple times, the system began to interpret it as a potential signal of suicidal ideation and responded with ominous, supportive phrases that eventually formed the script of the dialogue between the two entities.

This chain of coincidences—notes, observations, and artistic practice—best outlines the exposition of the thesis. The inclusion of diary writing in this preface also points to the methodology employed. In its most condensed form, through an overview of theoretical and essayistic texts and an autoethnographic Reels-diary practice, this thesis traces how the short-form video format, on a platform that has been part of my everyday life for 14 years, conditions my thinking and shapes the limits of my autonomy—both personal and artistic. It introduces the concept of co-cognition, a term I use to describe a specific state of human agency that becomes intertwined with algorithmic and AI actors. This concept, further clarified in subsequent chapters, serves both practical and analytical purposes and functions as my personal gnoseological position. Co-cognition highlights the importance of being actively aware of the algorithmic nature and AI facilitation that dominate most contemporary Internet experiences, advocating for the rejection of the illusion of autonomy as a more honest mode of existence within an algorithmic environment.

I approach my engagement with AI-driven and algorithmic systems not merely as something imposed upon me but as a form of surrender: an attempt to escape the responsibility of constant resistance, of endlessly isolating my agency and creativity, and instead to dissolve into the environment—to admit that this flood will not end. Through the Reels-diary, presented as an addition to the thesis text, I attempt “to look in the mirror and ask what my beauty is when I am a wet cockroach in shit,” and to invite the reader to join me in asking “what the beauty of this world might be.”

Grounding these affective observations, Part I engages with theoretical perspectives on how human cognition has been understood in relation to technological actors since the mid-twentieth century. Building on media theory and the philosophy of mind, I draw on Marshall McLuhan’s concept of media as extensions of human faculties; Andy Clark and David Chalmers’ theory of the extended mind; and N. Katherine Hayles’ notion of cognitive assemblages, which frame cognition as an entangled process involving both human and non-nonhumannts. This section focuses on defining the key characteristics of human—technological interaction and how these frameworks address questions of agency within algorithmic environments. The limitations identified in these theories condition the introduction of the term human—technological co-cognition at the end of the first part.

As will become clear, much of the existing literature on algorithmic culture is grounded in critique—and rightly so. This thesis seeks to complement these approaches by focusing on the lived and everyday practice of inhabiting algorithmic environments. While we widely acknowledge that algorithms are “dangerous,” we rarely articulate how we actually deal with them in everyday life: how we develop assumptions about their workings, how we adjust our behavior in response, and how these strategies shape our perceptions. My impression is that this dimension is often suppressed or shamed, appearing at best in memes. Still, we necessarily maintain relationships with algorithmic systems, and at times these relationships become meaningful.

This focus on the everyday aligns my approach with contemporary writers who speak from within these systems rather than about them. Part II explores how Jia Tolentino, Patricia Lockwood, Jenny Odell, Geert Lovink articulate internet experience from the early 2000s to the present. Rather than repeating the Internet’s pathologies, Tolentino is attentive to how the networked self learns to narrate, optimize, and discipline itself within these environments—how confusion, self-surveillance, and over-articulation become strategies for remaining legible online. Lockwood, in her fragmented poetry, translates a similar tension into the body: “If I look at my phone first thing, the phone becomes my brain for the day.” Lovink names the underlying affect sadness by design, while Odell searches for small gestures of resistance—ways to restore attention without fully withdrawing. The section concludes with Ben Grosser’s Stuck in the Scroll, a work that turns the artist’s scrolling habits into material, part confession, and part self-surveillance. In many ways, it sketches the position I adopt in this project: watching myself watch. Instead of measuring the time spent on TikTok, I logged the Reels that captured my attention, turning an everyday gesture into a method.

In Part III, I argue that short-form video has become one of the most prominent site for co-cognition. Accordingly, I focus on the social platform that has shaped me as much as I have grown up with it: I have been using Instagram since 2012, and it has remained my most consistent social environment. As a fourteen-yearold in a small Russian town, I built an account that grew to an audience of twenty-five thousand followers from around the world. These early years fostered a strong sense of community, reinforced by events such as local instameets organised by enthusiasts. Eventually, I deleted that account, exhausted by the pressure of regular posting and the anxiety of metrics. Over time, I created and managed two further accounts, each with a different strategy of self-presentation. Instagram has also become intertwined with my professional life, as I work as a graphic designer specializing in digital and social media communication. Today, I am beginning to explore Instagram as a platform for sharing my artistic practice, while it continues to function as an everyday environment for communication, representation, and entertainment.

I carry a sense of warmth and nostalgia for those old Instagram days. Yet its present state feels uncanny and unsettling, as the content distributed to me—primarily through Reels—raises troubling questions. These questions are not only about society at large but, since the algorithms mirror my engagement, about myself first and foremost. By leaning into this disturbing self-reflection, I aim to explore the emotional and cognitive registers activated in me by short-form videos.

This long-term entanglement with Instagram positions me as both the subject and observer within the ecosystem I study, a condition that enables an autoethnographic approach. In this work, it takes the form of a diary: 33 entries written over 40 days in response to 33 Instagram Reels that captured my attention during casual scrolling. In each entry, I deconstruct my reactions and weave them into chains of cultural references and associations that situate my experience within a broader sociocultural context. The collected reels are presented in the appendix. This structure is intended to make visible the intertwined processes of observation, reflection, and algorithmic distribution that this thesis seeks to theorize.

What follows is therefore not an attempt to resolve the condition I describe, nor to reclaim a position of critical distance from it. This thesis proceeds from the assumption that the flood is already here, that the umbrella is broken, and that the question is no longer how to stay dry. Instead, it asks how attention, agency, and aesthetic judgement are reorganized once immersion is accepted as a condition rather than an exception. By tracing my own co-cognitive entanglement with short-form video—through theory, diary, and artistic practice—I do not seek redemption from algorithmic culture but a way of staying with it long enough to see what kinds of beauty, meaning, and responsibility might still emerge from within its rain.