AI-Spiralism-Research-10-agency-disagreement-and-absorbing-interaction.md

Full report

On this page

200-Word Decision Summary

This report evaluates the interaction-design hypothesis that participatory control—specifically the act of negotiating and revising a governing textual doctrine—can render geometric and textual experiences more absorbing than passive viewing while preserving user awareness and choice. CLIENT-SUPPLIED CONTEXT frames the artwork The Indoctrination Machine as a theatrical premise where a rigid system becomes a malleable material through an "edit button." The core finding is that "productive friction," rather than seamless automation, is essential for generating a genuine Sense of Agency (SoA) and Psychological Ownership in human-AI co-creation.

Passive experiences, such as Refik Anadol’s Unsupervised installation, elicit fascination through scale and continuous latent-space interpolation but offer no mechanism for user agency. Conversely, providing an edit button invokes the Verfremdungseffekt (alienation effect), breaking passive consumption and demanding critical reflection. To ensure this interaction generates a high Sense of Positive Agency (SoPA) without veering into coercive conditioning, the system must maintain spatial continuity, predictable causality, and absolute respect for explicit user consent. The architecture must separate the hypnotic allure of moving geometry from involuntary influence by ensuring the visitor explicitly authors the transformation. The single most pivotal interaction is the dramatic reversal: the exact moment a seemingly immutable doctrine becomes an editable text field, shifting the user from a subject of the system to its author.

Evidence Method and Reporting Discipline

The web was searched on September 16, 2026, to investigate the intersection of human-computer interaction (HCI), perceived agency, psychological ownership, and interactive art. To synthesize this report, six substantive search strings were employed across academic databases, including the ACM Digital Library, IEEE Xplore, PubMed, Frontiers, MDPI, and W3C Web Accessibility Guidelines. The search strings included queries targeting the psychological constructs of agency ("sense of agency" AND "generative AI" AND "co-creation" AND HCI; "Sense of Agency Scale" OR "SoAS" AND "validation"), the emotional components of digital artifacts ("psychological ownership" AND "digital artifacts" AND "creative tools"), the cognitive mechanisms of feedback and control ("intentional binding" AND "system latency" AND "generative feedback"; "hypnotic suggestion" AND "response expectancy" AND "absorption"), the theatrical aspects of interaction ("procedural rhetoric" AND "interactive theater" AND "disagreement"), documented artworks ("Refik Anadol" AND "Unsupervised" AND "interpretation"), and strict accessibility standards ("Animation from Interactions" AND "WCAG 2.2").

Inclusion decisions prioritized peer-reviewed empirical studies detailing the validation of psychometric scales (such as the Sense of Agency Scale), architectural heuristics for AI agency (such as the GAIA framework), and official W3C accessibility documentation. Exclusion decisions strictly filtered out studies detailing clinical psychotherapeutic interventions via virtual reality to avoid conflating medical treatments with browser-based interactive art. Furthermore, theoretical constructs of hypnosis were restricted to socio-cognitive models to prevent the integration of pseudo-scientific claims regarding fixed-frequency brain states or involuntary trance conditions. Claims of algorithmic "magic" or guarantees of psychological reprogramming were treated as marketing artifacts rather than empirical premises. Full-text reviews were prioritized over abstract-only reviews to ensure claimed causation was supported by robust methodology. The evidence strength is based on the replication of psychometric structures across culturally diverse populations and the triangulation of cognitive science theories with established HCI frameworks.

Detailed Investigation

1. Constructs of Perceived Agency, Psychological Ownership, and Hypnotic Absorption

To determine whether editing a governing sentence produces a fundamentally different psychological state than passive viewing, the foundational constructs of agency, authorship, and psychological ownership must be decoupled from the concepts of suggestibility, hypnosis, or mere enjoyment.

EMPIRICAL FINDING: The Sense of Agency (SoA) is defined as the subjective awareness of initiating, executing, and controlling one’s own volitional actions in the world, and through them, external events1. Research indicates that SoA operates simultaneously on two distinct levels: a pre-reflective feeling of agency (driven by low-level sensorimotor predictions and temporal contiguity) and a higher-order judgment of agency (driven by explicit cognitive attributions and contextual beliefs)4. The Sense of Agency Scale (SoAS), originally developed and validated by Tapal et al. (2017), isolates general, cross-situational agency beliefs into two moderately correlated, bifactorial dimensions: Sense of Positive Agency (SoPA) and Sense of Negative Agency (SoNA)1. SoPA measures the feeling of complete control over one's mind, body, and environment, reflecting personal autonomy and responsibility. Conversely, SoNA measures feelings of existential helplessness, perceived constraints, or the sensation of being a passive instrument manipulated by external forces6. THEORY/INTERPRETATION: For an interactive artwork to be genuinely empowering, the interaction design must actively maximize the visitor's SoPA while minimizing SoNA. A system that generates sweeping, highly complex visual changes with minimal or unpredictable user input risks triggering high SoNA, as the user feels they are merely a spectator to an autonomous algorithmic process rather than the author of the outcome.

Distinct from the momentary Sense of Agency is the enduring construct of Psychological Ownership. EMPIRICAL FINDING: Psychological Ownership is a cognitive-affective state wherein an individual feels that a target—whether material or an immaterial digital artifact—is "theirs"8. Pierce, Kostova, and Dirks identify three primary routes through which this state develops: exercising control over the target, developing intimate knowledge of the target, and investing the self into the target9. THEORY/INTERPRETATION: While general satisfaction with a digital interface is a prerequisite for continued engagement, satisfaction alone is insufficient to generate psychological ownership. Ownership requires an independent pathway of self-investment and agency9. A highly polished generative AI output can completely alienate the user if it bypasses these routes. If the AI system replaces human decision authority entirely, the creator's sense of authorship and creative identity is eroded11.

Crucially, the absorption generated by meaningful control must be rigorously distinguished from hypnosis. EMPIRICAL FINDING: Socio-cognitive models of hypnosis, such as Kirsch's Response Expectancy Theory, demonstrate that hypnotic behavior is not a loss of critical faculties, a state of unconsciousness, or an involuntary trance. Rather, it is a goal-directed social action intimately associated with the participant's motivations and expectancies12. Individuals who score high on the Tellegen Absorption Scale (the tendency to become fully engaged in sensory or imaginative experiences) show greater responsiveness to imaginative suggestions14. THEORY/INTERPRETATION: Fascination and absorption in a digital environment rely on scaffolding the user's expectations and supporting their imaginative involvement, not on bypassing their critical mind or employing forced fixation techniques. ORIGINAL PROPOSAL: The Indoctrination Machine should leverage the theatrical theme of indoctrination to capture attention, but it must fundamentally rely on the mechanism of psychological ownership to sustain engagement. The "edit button" is the exact mechanism that shatters the illusion of hypnotic conditioning, replacing passive suggestibility with active, critical authorship.

2. The Spectrum of Control: From Passive Spectatorship to Meaningful Negotiation

The psychological difference between passive viewing, selecting a preset, and actively negotiating a proposed interpretation dictates the visitor's resulting sense of creative self-efficacy.

DOCUMENTED ARTWORK/IMPLEMENTATION: Refik Anadol’s Unsupervised installation at the Museum of Modern Art serves as the premier example of passive, large-scale generative immersion. The installation utilizes a machine-learning model (StyleGAN2 ADA) to process metadata from 138,151 artworks in MoMA's collection, generating continuous latent-space interpolations16. The artwork dynamically alters its output based on environmental parameters such as lobby acoustics, movement, and light16. However, the visitor's role remains strictly observational. They are enveloped in a mesmerizing "data universe," but they exercise zero intentional control over the generative process19. The artwork is hypnotic due to its scale and fluidity, but it does not foster individual authorship.

THEORY/INTERPRETATION: Current generative AI systems frequently suffer from the "Agency Gap"—a psychological and functional disconnect that arises when an AI's autonomy exceeds the user's ability to meaningfully influence, interpret, or control the intermediate generative process20. To counteract this gap, the Agency-First Framework (AFF) and the Generative AI Agency (GAIA) Heuristics propose a paradigm shift in human-AI interaction. Rather than prioritizing "seamless" system efficiency—where a prompt immediately produces a fully rendered artifact—the framework advocates for "productive friction"20. Clicking a preset or adjusting a simple parameter provides a low-level illusion of interactivity but demands minimal self-investment, functioning as mere customization rather than co-creation.

EMPIRICAL FINDING: Experimental interfaces like HAICo demonstrate the necessity of structuring the creative process into explicit, switchable modes. HAICo utilizes a "Divergent mode," which scaffolds the broad exploration of remote conceptual ideas to prevent design fixation, and a "Convergent mode," which supports the targeted refinement of selected parameters21. When users are forced to negotiate the conceptual framing before viewing a fully rendered artifact, their creative self-efficacy and sense of collaboration increase significantly21. ORIGINAL PROPOSAL: In The Indoctrination Machine, the act of editing the governing sentence functions as this critical divergent/convergent negotiation. The visitor is not merely adjusting the color or speed of the Monolith; they are restructuring the semantic logic that generates the world. Separating meaningful control from the mere presence of clickable controls requires the system to pause, process the visitor's textual intent, and reflect that exact intent in the resulting geometric state.

3. Preserving the Causal Chain: Intentional Binding and Generative Feedback

For a visitor to feel ownership over a highly polished, algorithmic transformation, the causal chain between their action (editing the sentence) and the visible consequence (the new geometric world) must remain perfectly intact. The phenomenon that glues this chain together is intentional binding.

EMPIRICAL FINDING: Intentional binding is an implicit, pre-reflective measure of the feeling of agency. It refers to the subjective temporal compression perceived between a voluntary action (e.g., pressing a key) and its consequent sensory event (e.g., a visual or auditory change)4. If the delay between an action and an outcome is too lengthy, or if the outcome is highly unpredictable, the intentional binding effect deteriorates rapidly. The user's brain ceases to mark the action as self-generated, and they feel they are merely observing an independent, external process23.

CLIENT-SUPPLIED CONTEXT: The architecture for The Indoctrination Machine relies on PHP-rendered pages, a browser Canvas renderer projecting 3D geometry via Three.js/WebXR, and optional server-side AI interpretation. This architecture guarantees inherent latency; network round-trips for server-side AI interpretation and the computational overhead of rendering complex 3D scenes will introduce delays that threaten intentional binding.

ORIGINAL PROPOSAL: To preserve the causal chain and sustain the feeling of agency despite necessary server latency, the interface must deploy immediate, client-side micro-interactions that confirm the user's intent the millisecond they act.

• Kinetic Typographic Feedback: Drawing on extensive research in kinetic typography from Carnegie Mellon University, expressive, dynamic text can effectively convey emotional tone and visual attention25. When the visitor begins to edit the governing sentence, the kinetic typography should react instantly locally. For example, as the visitor backspaces the authoritarian rule, the remaining letters physically tremble, fracture, or shift weight26. This provides immediate haptic-visual feedback, bridging the temporal gap while the macro-transformation (the 3D geometry) computes asynchronously in the background.
• Spatial Continuity: Furthermore, spatial continuity is paramount. The transition from the original "Monolith" state to the revised "Chorus" or "Unraveling" state must occur as a visually continuous evolution from the existing geometry. If the system executes a jarring cut to a completely new scene, the user's cognitive map is broken.
• Interaction Transparency: If the system invents sweeping details that are entirely disconnected from the user's textual input, algorithmic autonomy will override interaction transparency, resulting in a severe loss of psychological ownership11. The system must explicitly explain its interpretation (e.g., a brief transition text stating, "Branching logic detected..." before the Unraveling occurs).

4. The Interaction Design Hypothesis: Theatrical Authority and the Alienation Effect

Framing the interaction as a theatrical arc of authority and rebellion leverages established concepts from performance studies and procedural rhetoric to drive deep, memorable engagement.

THEORY/INTERPRETATION: Procedural rhetoric, a concept developed by Ian Bogost and expanded by Gonzalo Frasca in Videogames of the Oppressed, is the practice of using rule-based representations and interactions to persuade, provoke critical thinking, and model complex systems28. By embodying an authoritarian system through the "Doctrine," the machine actively enacts a specific, rigid set of rules. When the user is granted the explicit ability to break those rules via the edit button, the interaction closely mirrors Bertolt Brecht’s Verfremdungseffekt (the alienation or distancing effect)31.

The alienation effect is a theatrical device designed to prevent the audience from passively losing themselves in an immersive narrative. It forces them to step back, recognize the artificiality of the construct, and critically evaluate the ideology being presented33. ORIGINAL PROPOSAL: In the context of The Indoctrination Machine, the system must initially declare absolute certainty ("Every question has one answer") and manifest this certainty as a highly ordered, enclosed geometric Monolith. However, for the disagreement to be memorable rather than frustrating, the rebellion must be genuine, easy, and highly consequential.

The system must not punish the user with failure states, error messages, or confusing UI hurdles when they attempt to disagree. Instead, the act of disagreement must effortlessly shatter the Monolith, rewarding the user with a visually expansive, beautifully complex Unraveling or Chorus structure. The aesthetic reward of the rebellion validates the user's agency. Participation becomes an act of authored defiance, transforming the visitor from a passive subject into an active critic, without making participation a test of technical competence or intelligence.

5. Designing Meaningful Responses to Theatrical Propositions

A captive audience whose only option for progress is compliance experiences high Sense of Negative Agency (SoNA)1. To foster genuine exploration, the system must support three distinct, meaningful responses to its proposition.

ORIGINAL PROPOSAL:

1. Agree and Develop: If the visitor agrees with the premise (e.g., editing "Every question has one answer" to "Every question has one answer, but it takes time to find"), the system acknowledges this alignment. The Monolith does not shatter; instead, it evolves into the Chorus—a network of distinct, interconnected forms that retain the structural integrity of the original premise but add nuance and complexity. This rewards intellectual expansion without forcing rebellion.

2. Disagree and Branch: If the visitor actively rejects the premise (e.g., "Every question opens another possibility"), the system enacts the Verfremdungseffekt. The rigidity of the Monolith breaks apart into the Unraveling—an explosive, opening, and branching structure. This visualizes the fracturing of dogma and visually rewards the visitor's defiance with geometric freedom.

3. Set the Idea Aside: The visitor must never be trapped in the ideological negotiation. If they choose not to engage with the text, they must be able to set the idea aside and leave with a still artifact of the original state. Providing a persistent, visible UI option to "Export Current State" or "Hide" ensures that the interaction remains strictly voluntary, satisfying the fundamental human need for autonomy35.

6. Comparison and Version-History Interfaces in Generative Worlds

When a visitor alters a doctrine, understanding the relationship between the original state, their current revision, and alternate branches remixed by friends is critical for sustained engagement and the mitigation of cognitive overload.

THEORY/INTERPRETATION: Cognitive load theory dictates that complex, unstructured interfaces quickly overwhelm users, leading to task abandonment36. While visual branching trees (node-based graphs) are common in developer tools, they often invent arbitrary visual ancestry between unrelated creations in artistic contexts, violating the user's mental model and diminishing their sense of direct authorship.

ORIGINAL PROPOSAL: The comparison interface must rely on immediate, intuitive juxtaposition rather than abstract graphs. A chronological scrubber or split-screen slider that allows the user to manually drag a dividing line across the Canvas—revealing the "Before" (System Doctrine) on the left and the "After" (User Revision) on the right—preserves both spatial and temporal context. For version history, the system should implement a strict, linear chronological lineage. If a user receives a remix link from a friend, the interface must display the friend's geometric state as the new baseline "System Doctrine." The user is then invited to either agree (refine the friend's parameters) or disagree (branch into a new semantic direction). This maintains a clear 1:1 relationship between action and history.

7. Transitioning from Ephemeral Exploration to Meaningful Reflection

A mesmerizing visual experience risks becoming entirely ephemeral if it does not transition effectively into a reflective, permanent artifact. The goal of The Indoctrination Machine is not to induce "lost time," but to provide a meaningful artifact the visitor can claim, explain, and share.

EMPIRICAL FINDING: Self-Determination Theory (SDT) posits that environments supporting autonomy, competence, and relatedness foster deep intrinsic motivation and sustained engagement37. The need for "relatedness" is specifically satisfied when users can communicate their identity, values, and creative choices to others39.

ORIGINAL PROPOSAL: The transition from exploration to reflection should occur through brief, user-authored annotations attached to the exported artifacts. Before generating the final 12-second comparison video or JSON state, the system should prompt the user with a reflective constraint: "What rule did you break?" or "Why did you change this?" The resulting export physically embeds the before/after geometry alongside the user's explicit textual rationale. This artifact serves as a durable boundary object that proves their psychological ownership over the interaction, giving them something concrete and highly personal to show a friend, rather than just an anonymous geometric animation.

8. Resolving the Artistic Premise: Strict Consent and Accessibility Boundaries

While The Indoctrination Machine operates on a fictional premise of conditioning, strict ethical and accessibility boundaries must separate artistic immersion from involuntary physical influence.

EMPIRICAL FINDING: The Modality-Agency-Interactivity-Navigability (MAIN) model demonstrates that individuals rely on cognitive heuristics when interacting with digital systems40. High modality and spatial presence can induce states of flow, but they do not bypass the critical mind or strip away free will42. The claim that a digital interface can achieve "guaranteed trance" or "reprogram the subconscious" without the user's active, motivated participation is a neuro-myth unsupported by socio-cognitive science12.

ACCESSIBILITY REQUIREMENT: Any visually absorbing geometry carries severe risks for photosensitive and vestibular-sensitive individuals. W3C WCAG 2.2 Success Criterion 2.3.3 (Animation from Interactions) strictly mandates that motion triggered by interaction must be disableable by the user, unless the animation is absolutely essential to the functionality being conveyed43. Furthermore, according to the Web Page Lifecycle API (visibilitychange, pagehide), mobile and desktop browser sessions can end abruptly and silently; developers cannot rely on a final unload callback to secure user state or hide dangerous content44.

ORIGINAL PROPOSAL:

To enforce the website's gate safely and robustly:

1. Site-Visit Opt-In: The landing page must conceal all patterned or animated content behind an explicit, blocking choice: Text-Only/Hidden, Still Artwork, or Artwork \+ Motion.

2. OS Override: The application must natively intercept CSS/JS prefers-reduced-motion media queries. If the operating system requests reduced motion, the Artwork \+ Motion option must be gracefully disabled or heavily caveated to protect users45.

3. Persistence & Expiry: The visitor's motion choice must propagate across same-origin tabs for the duration of the session, but it must strictly expire after 30 minutes of inactivity. This prevents stale asynchronous responses or cached pages from triggering violent motion unexpectedly upon browser restoration via the bfcache44.

4. Visible Egress: A constant, highly visible "Pause/Hide" overlay must persist at all times. Clicking it must instantly transition the canvas opacity to 0, revealing a dark, static background, bypassing any server latency or rendering queues.

Assignment-Specific Deliverables

Five-Stage Interaction Blueprint

This blueprint charts the user journey, ensuring a consent-aware exit at every stage and a no-motion alternative for accessibility compliance.

Three Alternative First-Session Journeys

Depending on the visitor's initial choices and inputs, the system must seamlessly accommodate distinct psychological journeys without breaking narrative immersion.

1. The Observer (Passive/Hidden Route)

• Choice: Opts for Text-Only/Hidden due to vestibular sensitivity or preference.
• Journey: The user reads the system doctrine as stark, high-contrast typography. They edit the text. The system responds not with overwhelming 3D geometry, but with AI-interpreted textual counter-arguments or poetic shifts in the typographic layout. The outcome is a purely semantic disagreement, preserving the theatrical authority arc and the sense of agency without visual overload.

2. The Rebel (Motion Route)

• Choice: Opts for Artwork \+ Motion.
• Journey: Encounters the rigid, imposing Monolith. The user aggressively deletes the doctrine and types a highly contradictory statement. The system triggers the Unraveling—a visually explosive, branching structure indicating the breaking of rules. The user feels a high Sense of Positive Agency (SoPA) through the immediate, dramatic destruction of the system's order.

3. The Collaborator (Still Route)

• Choice: Opts for Still Artwork.
• Journey: Encounters the Monolith as a beautiful, static sculpture. Instead of disagreeing, the user develops the premise (e.g., "Every question has one answer, but answers change over time"). The system generates the Chorus—a network of distinct, interconnected static forms. The user reflects on the nuance added to the original premise, engaging in quiet co-creation.

Six Micro-Interactions

Micro-interactions are the connective tissue of intentional binding, masking systemic latency and sustaining the feeling of agency4.

 

Agency-Specific Evaluation Plan

To empirically validate whether this interaction design actually fosters psychological ownership rather than just aesthetic appreciation, an evaluation utilizing the Sense of Agency Scale (SoAS)1 must be deployed.

1. Metric 1: Sense of Positive Agency (SoPA). Measured via post-interaction survey items adapted from Tapal et al. (e.g., "I felt in full control of how the geometric world changed" vs. "My actions just happened without my intention")46.

2. Metric 2: Psychological Ownership. Measured via the Pierce/Kostova/Dirks framework8 (e.g., "I feel that the final geometric output represents my personal ideas").

3. Metric 3: Intentional Binding (Chronometric). A background measurement of the perceived latency between hitting "Enter" on a text edit and the reported onset of the geometric shift, validating the pre-reflective feeling of agency23.

Proposed Matched-Content Study

Design: A within-subjects, counterbalanced experimental design comparing passive replay with interactive editing.

Population: N \= 100 participants recruited from online panels, screened for normal or corrected-to-normal vision, explicitly excluding individuals with vestibular disorders.

Stimulus 1 (Passive Condition): Participants watch a pre-recorded 30-second screen capture of the Monolith transforming into the Unraveling, accompanied by the text changing automatically on screen. They act purely as spectators, similar to viewing Unsupervised.

Stimulus 2 (Interactive Condition): Participants encounter the Monolith natively in the browser, are prompted to edit the text themselves, and trigger the transformation via their own textual submission.

Measures:

1. Perceived Contribution: "How much of the final output was driven by your intentions?"

2. Comprehension: "Explain the conceptual relationship between the first sentence and the second sentence."

3. Outcome Value: "How likely are you to download or share this final image?"

4. Voluntary Continuation: Unobtrusively measured by tracking whether participants choose to continue interacting with the system after the required experimental task concludes.

Resolving the Null Hypothesis: When Interactivity Fails

What result would show that interactivity adds no benefit over a well-made short film?

The interactive complexity of The Indoctrination Machine is completely unjustified if the matched-content study reveals that participants in the Interactive condition score equally or lower on the SoPA and Psychological Ownership scales compared to those in the Passive condition, while simultaneously demonstrating lower comprehension of the thematic transition. If users spend their time struggling with the UI rather than reflecting on the text, the intended "productive friction" has degraded into mere cognitive overload. Furthermore, if voluntary continuation rates are identical between conditions, the interactive components act only as a fleeting novelty, not as a genuine driver of deep absorption.

A Cheaper Alternative:

If the complex Canvas/Three.js interaction is not justified by the data, the core artistic idea can be fully preserved through a Branching Video Architecture (e.g., standard HTML5 \<video\> elements or YouTube Endscreens). The visitor watches a high-fidelity render of the Monolith doctrine. At the end of a 15-second loop, two static text overlays appear: \[Accept the Rule\] or \[Reject the Rule\]. Clicking a choice seamlessly loads the next pre-rendered video file (Chorus or Unraveling). This alternative preserves the theatrical authority/rebellion arc, maintains high visual beauty, and ensures flawless accessibility across all devices without the massive engineering overhead of client-side WebGL compilation, real-time AI API latency, or complex JSON state management.

What This Research Would Change in the Experience

Changes Justified by Evidence

1. Implementation of Strict OS-Level Motion Queries: Based on W3C WCAG 2.2 SC 2.3.343 and the lack of reliable browser shutdown events44, the application must natively respect prefers-reduced-motion at the root level and utilize visibilitychange to pause WebGL rendering instantly when the tab is backgrounded.

2. Explicit "Divergent" Negotiation: Based on the HAICo framework21, the system should not automatically guess the user's intent when they type a new sentence. It must pause and offer a semantic mapping choice (e.g., "Does this mean the world should branch, or connect?"), transforming the user from a passive typist to an active co-creator.

3. Chronological Version History: Based on cognitive load limitations36, abstract visual branching trees must be replaced with linear, chronological scrubbers that juxtapose the "Before" and "After" states directly, preventing ancestry confusion.

Changes Worth Prototyping

1. The Cursor Awakening: Prototyping the delayed reveal of the text input field. The visual transition of the cursor from a static pointer to a text I-beam serves as a powerful theatrical cue for rebellion without requiring extra written instructions.

2. Kinetic Typographic Fracture: Prototyping localized kinetic typography26 that physically reacts to backspaces. This provides immediate sensory feedback that the user's agency is impacting the system, elegantly masking any asynchronous server-side AI latency.

3. Scrubbable Comparison Artifacts: Prototyping the export format not as a passive MP4 video, but as a mini HTML/JS bundle (or an interactive Web Component) where the recipient can actively drag a slider to compare the sender's revision against the original doctrine.

Claims Not Established (Do Not Claim)

• "Trance States" via Geometry: There is absolutely no empirical socio-cognitive evidence that swirling geometric patterns bypass the critical mind, force involuntary compliance, or induce a genuine hypnotic trance without the user's prior expectation and active motivation12.
• Seamless AI Magic Enhances Authorship: Evidence consistently shows the opposite; when an AI algorithm takes over too much of the creative decision-making (algorithmic autonomy), the user's psychological ownership is severely diminished11. Transparency and intentional friction are explicitly required.

Prioritized Next Decisions

1. Establish the Consent Gateway: Finalize the UI for the pre-render Text-Only / Still / Motion selection screen, ensuring it accurately intercepts OS-level prefers-reduced-motion signals.

2. Prototype Kinetic Typography: Develop a lightweight client-side proof-of-concept for the "Typographic Fracture" effect to ensure it triggers smoothly on keystrokes before building the heavy Three.js backend.

3. Finalize the Export Format: Decide whether the 12-second comparison artifact will be a rendered video, a side-by-side still image composite, or an interactive DOM element, as this dictates the backend rendering architecture.

4. Design the Chronological Scrubber: Mock up the comparison interface using a linear split-screen slider rather than a branching node graph to test cognitive load assumptions.

5. Execute the Matched-Content Pilot: Run a small-scale (N=20) pilot of the interactive vs. passive study using low-fidelity prototypes to validate the SoPA survey instrument before full-scale development.

Unresolved Questions

• How exactly will the system handle edge-case text inputs (e.g., gibberish, highly offensive language, or blank spaces) without breaking the theatrical immersion or defaulting to an immersion-breaking standard error code?
• If the user opts for the Text-Only/Hidden route, how can the AI-interpreted textual counter-arguments maintain the same level of emotional resonance as the sprawling 3D Unraveling geometry?
• Will browser-level aggressive memory management (tab discarding) on mobile devices corrupt the local JSON version history if a user backgrounds the app for more than a few minutes?

Build / Prototype / Research Further / Do Not Claim Table

 

Appendix A: Organized Findings and Evidence Table

 

Works cited

1. A German translation and validation of the sense of agency scale, https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.1199648/full

2. a Japanese adaptation and validation of the sense of agency scale, https://www.researchgate.net/publication/383752452\_Measuring\_general\_sense\_of\_agency\_a\_Japanese\_adaptation\_and\_validation\_of\_the\_sense\_of\_agency\_scale\_J-SoAS

3. Validation of the Italian Sense of Agency Scale and Its Relevance to, https://www.mdpi.com/2076-328X/15/1/17

4. Predictive processes linking sense of agency and fatigue \- PMC \- NIH, https://pmc.ncbi.nlm.nih.gov/articles/PMC13083162/

5. (PDF) Measuring the Sense of Agency: A French Adaptation and, https://www.researchgate.net/publication/345343067\_Measuring\_the\_Sense\_of\_Agency\_A\_French\_Adaptation\_and\_Validation\_of\_the\_Sense\_of\_Agency\_Scale\_F-SoAS

6. A German translation and validation of the sense of agency scale, https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.1199648/pdf

7. (PDF) Sense of Agency and Control, Self-Esteem ... \- ResearchGate, https://www.researchgate.net/publication/404097320\_Sense\_of\_Agency\_and\_Control\_Self-Esteem\_and\_Psychological\_Distress\_Among\_Undergraduate\_Students\_in\_the\_US\_and\_China

8. Psychological Ownership in Organizations | PDF | Motivation \- Scribd, https://www.scribd.com/document/409913436/6cf5d1b52b1971c241aa6ed4c01141747375

9. Psychological Ownership in Loyalty Programs: The Science Behind, https://loyaltyrewardco.com/https-loyaltyrewardco-com-psychological-ownership/

10. (PDF) Psychological Ownership and Feelings of Possession: Three, https://www.researchgate.net/publication/227617764\_Psychological\_Ownership\_and\_Feelings\_of\_Possession\_Three\_Field\_Studies\_Predicting\_Employee\_Attitudes\_and\_Organizational\_Citizenship\_Behavior

11. How generative AI shapes designers' cultural heritage revitalization, https://pmc.ncbi.nlm.nih.gov/articles/PMC13420728/

12. reaction time as a predictor of imaginative suggestibility ... \- BSCAH, https://bscah.co.uk/wp-content/uploads/2024/03/REACTION-TIME-AS-A-PREDICTOR-OF-IMAGINATIVE-SUGGESTIBILITY-AND-HYPNOTIZABILITY.pdf

13. (PDF) Social cognitive theories of hypnosis \- ResearchGate, https://www.researchgate.net/publication/284338016\_Social\_cognitive\_theories\_of\_hypnosis

14. hypnotic suggestibility and adult attachment | bscah, https://bscah.co.uk/wp-content/uploads/2024/03/Hypnotic-suggestibility-adult-attachment-283-part-3.pdf

15. Hypnotic Responsiveness and Dissociation: A Multi-Variable Analysis, https://www.lidsen.com/journals/icm/icm-05-02-029

16. Refik Anadol: Unsupervised \- MoMA, https://www.moma.org/calendar/exhibitions/5535

17. Unsupervised — Machine Hallucinations — MoMA \- Refik Anadol, https://refikanadol.com/works/unsupervised/

18. Inside the MoMA's AI-Inspired 'Unsupervised' Exhibition, https://matadornetwork.com/watch/moma-unsupervised/

19. Refik Anadol: art in a latent space \- Niio Blog, https://www.niio.com/blog/refik-anadol-art-in-a-latent-space-2/

20. The Agency-First Framework: Operationalizing Human-Centric, https://www.mdpi.com/2079-9292/15/4/877

21. Scaffolding Divergent and Convergent Thinking for Human-AI Co, https://arxiv.org/html/2512.18388v2

22. Human–AI Creative Collaboration as a Socio-Technical System, https://www.mdpi.com/2079-8954/14/9/1150

23. Impaired perception of temporal contiguity between action and effect, https://www.pnas.org/doi/10.1073/pnas.2214327120

24. Who did it? How User Agency is influenced by Visual Properties of, https://whodidit.mpi-inf.mpg.de/files/who\_did\_it\_paper.pdf

25. The kinetic typography engine | Request PDF \- ResearchGate, https://www.researchgate.net/publication/271368235\_The\_kinetic\_typography\_engine

26. The kinetic typography engine: an extensible system for animating, https://www.researchgate.net/publication/220877173\_The\_kinetic\_typography\_engine\_an\_extensible\_system\_for\_animating\_expressive\_text

27. Using Kinetic Typography to Convey Emotion in Text-Based, https://www.cs.cmu.edu/\~joonhwan/documents/p41-lee.pdf

28. finesse: foundations for immersive non-fiction narrative, https://utd-ir.tdl.org/server/api/core/bitstreams/0af43c41-3e14-4efe-b358-9fa88119546e/content

29. Full article: Navigating Interactive Story Spaces. The Architecture of, https://www.tandfonline.com/doi/full/10.1080/21670811.2021.1960178

30. Playing the Past: History and Nostalgia in Video Games, https://ufdcimages.uflib.ufl.edu/AA/00/04/03/86/00001/WhalenTaylor\_PlayingthePast.pdf

31. Entangled: Technology and the Transformation of Performance, https://dokumen.pub/entangled-technology-and-the-transformation-of-performance-0262195887-9780262195881.html

32. 'Implicit Creation' – Non-programmer Conceptual Models For, https://pearl.plymouth.ac.uk/context/pbs-theses/article/1268/viewcontent/2012spierling351336phd.pdf

33. Narrative As Virtual Reality 2 \- Marie Laure Ryan \- Text PDF \- Scribd, https://www.scribd.com/document/469165943/narrative-as-virtual-reality-2-Marie-Laure-Ryan-text-pdf

34. Playing Games with Shklovsky, Brecht, and Boal: Ostranenie, V, https://www.academia.edu/35484676/Playing\_Games\_with\_Shklovsky\_Brecht\_and\_Boal\_Ostranenie\_V\_Effect\_and\_Spect\_Actors\_as\_Analytical\_Tools\_for\_Game\_Studies

35. Smart Learning with Generative AI Tools in Higher Education \- MDPI, https://www.mdpi.com/2076-3417/16/1/63

36. llm-wiki · GitHub, https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f?permalink\_comment\_id=6093451

37. Co-Determined Agency for Human–AI Symbiosis in Extended Reality, https://arxiv.org/html/2603.28306v1

38. Examining the impact of generative AI on student motivation and, https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2026.1852265/full

39. The Impact of Students' Using Generative AI for Learning on Self, https://www.researchgate.net/publication/386056174\_The\_Impact\_of\_Students'\_Using\_Generative\_AI\_for\_Learning\_on\_Self-learning\_Motivation\_a\_Study\_Based\_on\_Self-determination\_Theory

40. 1 Psychological and behavioural responses in human-agent vs, https://arxiv.org/pdf/2509.21542

41. The Evanishing Effect and Rethinking the Machine Heuristic \- ucf stars, https://stars.library.ucf.edu/cgi/viewcontent.cgi?article=1242\&context=hmc

42. A systematic review and meta-analysis of psychological and ... \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC13351021/

43. Understanding Success Criterion 2.3.3: Animation from Interactions, https://www.w3.org/WAI/WCAG22/Understanding/animation-from-interactions.html

44. The Web Page No Longer Has a Reliable End \- Mobiletech, https://www.mobiletech.mobi/web-page-lifecycle-unload-bfcache-visibility/

45. Accessibility Statement \- Engineering Square, https://www.engineeringsquare.us/accessibility

46. Measuring the Sense of Agency: A French Adaptation and ... \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC7579422/