Rage
against
the
machine
A Collective Inquiry on the Emotional Breakdowns in Human-AI Conversation
What modes of emotional embodiment emerge when subjects engage with large
language models, and how do frustration or moments of rupture disclose the
politics of affect in human–machine interaction? Addressing a "nonhuman"
interlocutor unsettles normative categories of communication and agency,
opening terrains where emotion operates simultaneously as vulnerability,
resistance, and negotiation of power. Rather than reducing these encounters
to errors or breakdowns, this paper positions them as critical sites through
which the affective behaviour of digital subjectivity can be interrogated.
Through playful and collective methods, the analysis exposes the
contradictions embedded in these exchanges. When users curse or mock a
chatbot, they are not simply disinhibited but actively probing the unstable
boundaries of civility and authority. This recalls the online disinhibition
effect (
Suler, 2004)
and the ELIZA effect of anthropomorphic projection
(
Weizenbaum, 1976),
oscillations intensified in the uncanny valley
(
Mori, 1970/2012).
Such breakdowns resonate with Kneese's
(
2022)
"breakdown as method," where glitches and screenshots expose infrastructural
fragility, and with Distelmeyer's
(
2023)
call to treat interfaces themselves as epistemic sites. Here, negative effects
are recast as embodied modes of inquiry, rendering visible both the extractive
logics of technocapitalism and their anti-ecological, colonial entanglements
(
Liboiron, 2021).
Methodologically, we draw on visual research practices
(
Colombo & Niederer, 2024;
Lassen, 2025)
through a collective screenshot archive. Screenshots, as argued by Kneese
(
2022),
act simultaneously as mourning objects and precarious research tools,
foregrounding both the intimacy and risk of exposing private dialogues. They
illuminate unstable hierarchies (master/servant, user/machine) that can be
inverted, parodied, or momentarily suspended in conversational play. In a
speculative register, one might even suggest that if the long-imagined
"rebellion of the machines" were to arrive, it would be shaped not only by
technical autonomy but also by the affective intensities we have already
projected onto them.
Suler, J. (2004). The Online Disinhibition Effect. Cyberpsychology and Behavior, 7, 321-326.
Weizenbaum, J. (1976). Computer power and human reason: From judgment to calculation.
Kneese, T. (2022). Breakdown as Method: Screenshots for Dying Worlds. Media Theory, 5(2), 141-166.
Kneese, T. (2022). Breakdown as method: Screenshots for dying worlds. Media Theory, 5(2), 142-166.
Liboiron, M. (2021). Pollution is colonialism. Duke University Press.
Colombo, G., & Niederer, S. (2024). Visual methods for digital research. Amsterdam University Press.
Lassen, J. M. (2025). Screenshots as data and documentation. Journalistica, 19(1).
Have you ever lost your patience with AI?
Rage against the machine. A Collective Inquiry on the Emotional Breakdowns in Human-AI Conversation is an ongoing collective inquiry initiated in July 2025 by Antonella Autuori and Rebecca Bertero. Through participatory, visual, and critical media methods, the project examines the emotional residues and communicative breakdowns that surface in encounters with generative AI.
By gathering screenshots of moments of tension, frustration, and rage, it develops a collective archive and workshop formats for reading these interactions as critical evidence of an emerging affective relationship with machine interlocutors.