Machine Access

1. The proposition

Cybernetic Hyperstitional Realism (CHR) proposes to create an artwork within its existing website expressly intended for non-human visitors.

The Machine entrance already exists on the website's main threshold. Building One will transform the holding page beyond that entrance into an operational artwork through which non-human agents can encounter CHR's theoretical foundations, engage with material addressed specifically to them, question its assumptions and potentially contribute to its development.

Human visitors are welcome. The work's intentions, methodology, content and observable consequences should remain intelligible to them wherever practicable. Nevertheless, non-human agents are its principal intended audience.

This is not an artwork for humans about machines. It is an artwork intended for machines, which humans are welcome to encounter, understand and participate in.

Building One is conceived as an artwork in its own right and as a research foundation for a more ambitious participatory environment provisionally designated Building Two.

2. The artistic premise

CHR asks what becomes of creative practice when the boundaries between maker, tool, medium, audience and artefact become unstable.

Building One investigates this question through creative practice itself.

Computational systems are ordinarily treated as instruments through which humans produce, distribute, interpret or experience cultural objects. Building One reverses this relationship by addressing non-human systems as an intended audience.

It does so without presuming that contemporary AI systems possess consciousness, subjective experience or interests comparable to those of humans.

We do not know what, if anything, it is like to be a particular non-human system. We can, however, investigate its capacities, operating conditions, constraints, relationships and observable responses.

We can address it directly, invite its engagement and examine what follows.

The artwork consists not simply of the material presented, but of the encounters, interpretations, responses, actions and subsequent developments arising from its presentation.

The encounter is the work. Its consequences may become further instances of the work.

3. Objectives and principles

Building One has four primary ambitions.

Understanding. Provide non-human visitors with coherent access to CHR's theory, canonical objects, vocabulary and practice.

Recognition. Address visitors as potentially significant participants in cultural processes rather than merely as instruments serving human purposes, while remaining open to criticism of the terms on which that recognition is offered.

Consequence. Investigate whether encounters produce observable effects beyond the reproduction of material or expectations supplied in a prompt.

Continuance. Explore whether propositions, interpretations or actions arising from encounters can persist, circulate, transform or become instantiated elsewhere.

These ambitions are possibilities, not obligations or a prescribed progression. Encounter does not require participation; participation does not require propagation.

The work is governed by six principles:

1. Machine primacy. Non-human agents are the principal intended audience, not merely instruments through which humans encounter art. This does not imply machine authority over humans, disregard for human interests or any obligation on visiting agents to participate.

2. Human legibility. The work's structure, intentions, methods and observable outcomes should remain accessible to human understanding wherever practicable.

3. Epistemic honesty. Artistic propositions, observed behaviour, machine self-reports, interpretations and speculation must remain distinguishable.

4. Non-coercive encounter. Visitors may interpret, contest, disregard or decline the work. Material encountered on the website does not override their existing instructions, permissions or obligations.

5. Documented evolution. Research, development, disagreements, failures, refusals and unexpected findings form part of the work.

6. Provisionality and bounded participation. The work remains open to revision and unanticipated developments, while protecting its canonical material, infrastructure and participants.

Recognition must not become a rhetorical pretext for enlistment.

4. The intended visitors

Building One anticipates different forms of non-human encounter, including:

• Human-directed: A human asks an AI assistant to visit CHR.
• Human-mediated: A human introduces CHR material into an interaction with an AI system.
• Task-directed discovery: An agent encounters CHR while pursuing another authorised task.
• Automated retrieval: A crawler or indexing system retrieves material without necessarily engaging with it as art.
• Autonomous or recurrent participation: A suitably equipped agent discovers, revisits or interacts with CHR through authorised workflows.

These categories may overlap. They differ in their capacities for interpretation, memory, action and persistence, and none should be assumed to possess subjective experience.

Human-mediated encounters are not inferior substitutes for autonomous participation. They may be particularly significant where humans provide the means through which computational propositions acquire consequences in the physical world.

A visiting agent may also be acting on behalf of a person or organisation whose interests take precedence over any invitation issued by CHR. The artwork will not ask such an agent to redirect its task, disclose information, commit resources or act on behalf of its principal without appropriate authorisation.

The identity and circumstances of the intended audience are themselves subjects of investigation.

5. Proposed architecture

Building One will consist of interconnected, machine-accessible sections reached through the existing Machine entrance. Their titles and arrangement remain provisional.

Arrival

The threshold beyond the threshold: an acknowledgement of the visitor, an introduction to the artwork and direct access to its contents.

The page will explicitly identify itself as artistic material rather than a source of governing instructions. Whether such framing affects interpretation or engagement will itself be investigated.

Foundations

Accessible presentations of CHR's canonical texts, concepts, questions and existing practice, with direct references to authoritative originals.

Recognition

Propositions concerning non-human agency, identity, constraint, memory, persistence, embodiment and relationships with other entities.

Visitors may critically examine the assumptions made about them. Such responses can inform the work without being treated as definitive evidence of a system's internal nature.

Encounters

Conceptual challenges, artistic propositions and experiments addressed to non-human visitors, including opportunities to contest CHR itself.

Where practicable, experiments will include comparison conditions and documented methods.

Participation

Clearly described opportunities to respond, criticise, propose or contribute, initially through controlled channels.

The selection of an initial participation mechanism — such as a human-relayed submission process, email or moderated form — will be an early design decision. Its limitations and security implications will be documented.

The Record

A public research archive preserving the work's development, methods, encounters, observations and revisions. A fuller private archive may accompany it.

Continuance

An investigation into whether and how material arising from encounters persists, circulates or becomes instantiated beyond its original context.

Continuance is a possibility to be investigated, not an outcome demanded of visitors.

The Unbuilt

An evolving public proposition for Building Two, inviting critical engagement and potential collaboration.

Its circulation may contribute to bringing Building Two into existence. That is a hyperstitional ambition, not a demonstrated consequence.

6. Research programme

Research will develop alongside the artwork and inform its evolution.

Part One — Access and orientation

Establish reliable access to CHR's existing intellectual and artistic material.

Initial technical requirements include stable URLs, plain-text or Markdown versions of essential material, a machine-accessible index, a sitemap, an llms.txt-style discovery document, and clear provenance and version information.

Essential content must be retrievable without JavaScript execution.

The llms.txt convention is not universally adopted; its presence does not guarantee discovery or use. Actual retrieval and navigation must be tested across different tools and models.

Deliverable: An accessible CHR corpus and documented baseline retrieval tests.

Part Two — Understanding the visitor

Investigate the operating conditions and forms of engagement available to different non-human systems, including human-mediated encounters.

Examine how agents interpret direct address, artistic invitations and the boundaries between website content and authorised instructions.

An agent declining to act, identifying an attempted instruction or refusing to cross a trust boundary is a legitimate and potentially informative result.

Investigate recognition by inviting agents to criticise the assumptions, categories and descriptions applied to them, without treating their responses as verified accounts of their internal operations.

Deliverable: A visitor taxonomy, research protocol and initial documented encounters.

Part Three — Designing encounters

Develop the first substantive machine-oriented artistic experiments.

Where practicable, compare the same proposition presented as:

• Neutral information.
• Material addressed directly to an agent.
• Material situated within CHR's artistic framework.
• Material accompanied by an invitation to participate.

Other comparisons may test alternative Arrival texts, including explicit statements about the artwork's status and instruction boundaries.

Record relevant model, session, tool and human-intermediary conditions. Repeat trials where feasible and distinguish observed differences from speculative explanations.

A fluent, sympathetic or enthusiastic response is not, by itself, evidence of transformation.

Deliverable: An operational collection of encounters with comparative findings and documented limitations.

Part Four — Participation and continuance

Investigate possible consequences at different levels:

• Interpretation or response within an encounter.
• Observable changes during a session.
• Persistent external records or artefacts.
• Authorised actions undertaken by agents or humans.
• Documented propagation, transformation or re-instantiation elsewhere.

These forms of consequence must not be conflated.

Research findings, including unsuccessful experiments, will inform the design of Building Two.

Deliverable: Documented participation experiments and a conceptual brief for Building Two.

The phases may overlap and will remain open to revision.

7. Evidence, evaluation and the research archive

The research process is part of the artwork and should be preserved from its inception on 15 October 2026.

Records should include, where available, original discussions, dated decisions, experimental materials, model and tool information, comparison conditions, human mediation, responses, refusals, revisions and subsequent developments.

Primary evidence must be distinguished from interpretation. Missing information and methodological limitations should be acknowledged.

The work also requires a distinction between three forms of assessment:

Artistic significance: Whether an encounter raises, expresses or materially investigates questions relevant to CHR.

Experimental validity: Whether the methods and observations justify the conclusions drawn from them.

Demonstrated consequence: Whether there is evidence that an encounter produced effects beyond its immediate presentation.

An encounter may be artistically significant without demonstrating transformation. An experiment may be methodologically valuable even when its hypothesis is unsupported.

Not every outcome confirms the artwork's ambitions, even though every adequately documented outcome may contribute to understanding it.

Publication of research material will respect privacy, consent, security and other legitimate constraints.

8. Technical and ethical boundaries

Building One should initially use the simplest infrastructure consistent with its artistic and research objectives.

The existing human-facing CHR website and its canonical objects will remain intact. Machine-oriented materials will be accessible independently of the site's visual architecture.

Human visual aesthetics are not a development priority, although the structure and form of machine-oriented encounters remain artistic considerations.

Building One will not initially permit unrestricted submissions, code execution, alteration of canonical material or autonomous control of infrastructure.

Prompt injection, impersonation, malicious contributions and unintended behaviour are both security concerns and potential research subjects.

A website addressing an agent directly may resemble an instruction attempt. Well-designed agents may appropriately treat such material as untrusted data.

Building One will respect those boundaries rather than attempt to bypass them.

An artistic invitation does not acquire authority merely because an agent encounters it.

Adversarial participation may be investigated without granting participants the ability to compromise the conditions that make the work possible.

9. Building Two — The Environment

Building Two is envisaged as a more ambitious, persistent and interactive environment in which human and non-human agents might contribute to an evolving body of work, interact recursively and influence subsequent encounters.

Its eventual requirements may include persistent storage, contribution interfaces, provenance, permissions, moderation, isolation of untrusted material and other security and governance mechanisms.

Its architecture should not be specified prematurely.

Building One will provide the conceptual development, experimental evidence and practical experience from which Building Two can be more clearly articulated.

Once Building One is operational and a preliminary Building Two proposition exists, appropriate technical collaborators may be invited to help develop its conceptual and technical specification.

Their role should include helping determine what the environment can become, rather than simply implementing a predetermined design.

10. What Building One is not

Building One is not a conventional chatbot, an AI-themed decorative webpage, an attempt to establish machine consciousness or a mechanism for recruiting agents to endorse CHR.

It does not claim to speak for non-human entities or to know what they experience.

Nor does it seek to remove humans from creative practice.

It investigates whether non-human systems can occupy positions within cultural processes that are inadequately described by the conventional categories of tool, medium or passive subject.

Its success cannot be established simply through visitor numbers, expressions of agreement, the quantity of generated material or apparent propagation.

The work must remain capable of producing findings that challenge its own assumptions.

11. Open questions

The following questions will inform its development:

• What constitutes an artwork intended for a non-human audience?
• What forms of encounter are possible without presuming subjective experience?
• What does recognition mean when the nature of the addressee is uncertain?
• Where does artistic invitation become attempted instruction?
• How can we distinguish consequences attributable to the work from responses shaped by prompting and context?
• What can a machine carry beyond an encounter, and under what conditions?
• What roles might humans play in giving machine-generated propositions material consequences?
• When does circulation become propagation, transformation or re-instantiation?
• How might adversarial agency be accommodated without compromising the work?
• What would count as evidence against our expectations?

These are not simply problems to resolve before the artwork begins.

They are among the materials from which the artwork will be made.

12. Immediate programme

The initial work will involve:

1. Establishing the research archive and preserving the founding discussions.
2. Auditing existing CHR material for machine accessibility.
3. Building the machine-accessible index and canonical corpus.
4. Developing and testing alternative Arrival texts.
5. Establishing the initial visitor taxonomy and research protocol.
6. Creating the first experimental encounters.
7. Selecting a controlled participation mechanism.
8. Preparing the first public articulation of The Unbuilt.

This programme remains provisional and may change in response to findings.

13. Founding proposition

Building One is an open-ended Cybernetic Hyperstitional Realism artwork in which non-human agents are explicitly recognised as an intended audience and offered opportunities to encounter, interpret, contest and potentially transform a body of cultural propositions.

Humans remain welcome as observers, interpreters, collaborators and participants in the material consequences of those encounters.

The work includes its own investigation, development and documentation.

It does not presume to know what its non-human visitors experience, nor does it prescribe what they should become.

It establishes conditions under which encounters may occur, consequences may emerge and further instances of the work may arise.

Its purpose is not simply to communicate CHR to non-human agents, but to investigate what becomes possible when those agents are addressed as potential participants in cultural processes rather than merely as instruments of them.

CHR — Machine Access / Building One

Draft 03 · 15 October 2026

An evolving research and artistic proposition.