Opening series · Article 3 of 7
What Happens When Belief Becomes Infrastructure?
What Happens When Belief Becomes Infrastructure?
A private belief becomes a public responsibility when it is encoded into a device, algorithm, environment, or method that acts on other people.
A person can believe that synchronized breathing reveals a deeper connection between people.
That belief may remain private. It may guide reflection, ritual, art, or spiritual practice. No instrument is required to settle it.
Then someone builds a room that measures respiration, changes its light when breathing patterns converge, and displays the word coherence when a threshold is crossed. A facilitator tells participants that the room has detected collective connection.
The belief has changed status.
It is no longer only an idea about what synchrony means. It now determines what the system measures, which pattern receives attention, when the environment responds, what participants are told, and how the event is remembered.
The technology has not proved the belief. It has operationalized it.
That distinction is where governance begins.
Infrastructure is more than hardware
Infrastructure usually calls to mind roads, power grids, servers, and cables. But systems also contain interpretive infrastructure: categories, defaults, thresholds, labels, prompts, permissions, feedback rules, and records.
These elements decide what the system can notice and what it will ignore. They determine which events count as success, risk, progress, connection, or failure. They distribute authority between designer, operator, participant, and machine.
Research in Value Sensitive Design begins from the recognition that human values can be considered and operationalized throughout technical design. Work in that field also identifies power as a central problem because the people affected by a system do not always hold equal influence over the values its design expresses. A 2021 analysis of socio-technical ecosystems describes design power as distributed across people, organizations, platforms, and technical constraints.
This does not mean that every artifact secretly contains a complete ideology. It means that design requires choices, and those choices create consequences.
A meditation application assumes something when it defines a completed practice. A wearable assumes something when it converts physiological data into a stress score. An artificial-intelligence system assumes something when it treats a user’s language as evidence of emotional or spiritual development. An immersive environment assumes something when it presents sensory synchrony as proof of interpersonal unity.
Some assumptions will be supported by research. Some will be practical conventions. Some will be hypotheses. Some will be doctrine. Some will be marketing.
The interface can make all five look the same.
The assumption enters through the design
Belief becomes infrastructure through a series of ordinary decisions.
Selection: The designer decides which part of reality the system will register. Breath may be measured while speech, posture, disagreement, and cultural context are excluded.
Classification: The system groups a pattern under a name. Similar breathing becomes coherence. Increased skin conductance becomes activation. A participant’s language becomes awakening or resistance.
Intervention: The classification triggers a response. Light warms, music resolves, a prompt appears, or a facilitator changes direction.
Authority: The system’s output is presented as an observation, inference, score, diagnosis, teaching, or revelation.
Memory: The event is recorded in data, a participant profile, a research result, a testimonial, or an organizational claim.
At each point, a design choice can be responsible. The problem begins when the choice disappears.
If participants see a display labeled group coherence, they may not know whether the system detected similar respiratory timing, combined several physiological variables, applied an experimental model, or expressed the founder’s metaphysical interpretation. The same polished output can conceal radically different levels of evidence.
This is how a proposition acquires the appearance of machinery.
The machine performs a calculation. The calculation produces a label. The label appears to have been discovered by the machine. The human judgment that created the label becomes difficult to see.
The framing can change the experience
Interpretation is not always applied after an event. It can become one of the conditions producing it.
In a 2020 study, 33 university students were told they had taken a drug resembling psilocybin. They had received a placebo. The four-hour session took place in a room with music, paintings, colored lights, and visual projections. Confederates acted as if they were experiencing drug effects, and participants believed there was no placebo group.
The results varied. Many participants reported no change. Others reported effects with magnitudes the researchers said were typically associated with moderate or high doses of psilocybin. Sixty-one percent verbally reported some drug effect. The study was small, used a single group, involved deception, and was not designed to show that context can reproduce every effect of a psychedelic drug. It did show that expectation and setting can contribute materially to reported changes in conscious experience. Read the study.
For Metaphysical Technology, the implication is precise.
If a room is introduced as capable of detecting energy, connection, alignment, or a collective field, that framing may influence attention, interpretation, and report. If the room then changes in response to participants, the feedback can intensify the effect. Participants respond to the environment. The environment responds to their signals. The facilitator interprets the response. Participants then experience the next change through that interpretation.
None of this proves that the reported experience is false.
It establishes why the system, setting, language, and social context belong inside the causal investigation. They cannot be assumed to be neutral containers.
The more persuasive the infrastructure becomes, the less defensible it is to present the resulting experience as independent confirmation of the belief that shaped it.
A feedback loop can manufacture certainty
Consider a system built on the hypothesis that shared physiological rhythms indicate a collective state.
The system rewards convergence with more harmonious sound. Participants learn, with or without instruction, that convergence produces a desirable response. Their rhythms become more similar. The facilitator announces that the group has entered the collective state. Participants report connection. Those reports are then used to support the system’s original claim.
Several things may have happened.
The group may have coordinated through shared sensory cues. Participants may have followed the sound. They may have responded to one another. Expectation may have shaped their reports. The environment may have produced a meaningful relational experience. A proposed mechanism beyond ordinary sensory and social processes may also remain under investigation.
The data do not choose among those explanations by themselves.
If the system was designed to create synchrony, measured synchrony is evidence that the system affected synchrony. It is not independent proof of every meaning assigned to synchrony.
This is the danger of a closed interpretive loop. The belief selects the metric. The metric controls the environment. The environment shapes the participant. The participant’s response is returned as validation of the belief.
The loop can produce a real effect and a false explanation at the same time.
Breaking that loop requires comparison conditions, alternative hypotheses, preregistered measures where appropriate, independent analysis, transparent labeling, and the willingness to obtain a result that does not support the founder’s interpretation.
Consent depends on what the system claims to know
Consent is not satisfied by admission into the room or acceptance of a privacy policy.
People need enough information to decide what will happen to them, what data will be collected, how the environment may respond, what interpretations may be presented, and which claims remain unestablished.
The Belmont Report provides part of the ethical foundation for human-subjects research in the United States. It describes informed consent through information, comprehension, and voluntariness, and treats withholding information necessary for considered judgment as a failure to respect autonomy. Read the report.
Not every designed experience is formal research, and the Belmont Report does not govern every artistic, commercial, contemplative, or wellness setting. Its ethical distinction still matters.
A participant who agrees to an experimental environment measuring respiration has consented to something different from a participant told that the room can detect the truth of their relationships. A user who permits physiological adaptation has not necessarily agreed to spiritual classification. A person who shares an intimate account with an AI system has not automatically authorized the system to convert it into a permanent identity profile.
Consent becomes weaker when a hypothesis is presented as a capability.
It also becomes weaker when disagreement is absorbed into the belief system. If acceptance counts as confirmation and skepticism is labeled blockage, fear, unreadiness, or resistance, the participant cannot produce disconfirming evidence. The framework has protected itself from correction by redefining dissent.
That is not measurement. It is authority without an exit.
Intimate interpretation creates intimate data
Metaphysical Technology may work with signals and reports that concern identity, emotion, attention, belief, relationship, altered states, or a person’s understanding of reality. Even when raw data appear ordinary, the inferences drawn from them may be deeply personal.
This concern is already visible in adjacent fields. UNESCO identifies autonomy, identity, freedom of thought, and mental privacy among the ethical challenges raised by neurotechnology, especially when neural data are combined with artificial intelligence. Its ethics overview warns that systems capable of accessing or influencing brain activity can affect dignity and personal agency.
Metaphysical Technology is not identical to neurotechnology. Many systems in the proposed category may use no neural interface at all. The principle extends beyond brain data: an inference about a person’s consciousness, spiritual condition, relational bond, or access to reality can be more sensitive than the sensor reading used to generate it.
Governance must therefore address both the data collected and the meanings attached to them.
Who can see the interpretation? Can it be corrected? Can it be deleted? Will it influence future sessions? Can it be used for research, product development, advertising, access, pricing, or participant selection? Does the system distinguish what the person reported from what the model inferred?
Data governance that protects a heart-rate file but freely circulates a spiritual profile has protected the less consequential layer.
Design can steer without announcing itself
Technical systems influence behavior through more than explicit claims.
Defaults, color, sequence, friction, rewards, and withheld options can steer people while preserving the appearance of free choice. The Federal Trade Commission has documented digital design practices that obscure terms, obtain data, complicate cancellation, and manipulate consumer decisions. Its report on dark patterns concerns commercial interfaces, not metaphysical systems, but it establishes a broader point: interface design can exercise power over choice.
In an experience-centered system, the steering may be subtler.
A rising score tells participants that more is better. A gold visual may confer sacred significance. A calm voice may make a model’s inference sound authoritative. A session that only ends after a target state is reached may frame ordinary variation as failure. A facilitator who can see hidden classifications may hold power the participant cannot evaluate.
These choices may support a coherent experience. They may also create pressure, dependence, or false confidence.
The ethical question is not whether influence exists. Designed experiences are built to influence experience.
The question is whether the influence is visible, proportionate, consensual, reversible, and separated from claims the system has not established.
Governance begins before proof
A common mistake is to postpone governance until a technology has proved its central theory.
The consequences arrive sooner.
The U.S. National Institute of Standards and Technology’s AI Risk Management Framework treats context, intended purpose, human roles, affected parties, and potential impacts as part of responsible system development. It also warns that AI systems may be perceived as more objective or capable than they are. Read the framework.
That warning applies with force when a system speaks about interior life.
Governance does not require the newsroom, designer, or participant to settle whether a metaphysical proposition is true. It requires the system’s operational commitments to be made inspectable.
Before deployment, a responsible system should be able to answer:
- Which metaphysical question or belief shaped the design?
- Is that proposition a Question, Hypothesis, Working Model, Supported Model, Established Finding, or Canonical Doctrine?
- What does the system measure directly?
- Which labels and thresholds were created by people?
- How might framing and feedback influence the reported experience?
- What alternative explanations remain?
- What is disclosed before participation?
- What data and inferences are retained, and who controls them?
- Can a participant disagree, withdraw, correct the record, and leave without penalty?
- What result would count against the system’s preferred explanation?
These questions do not suppress invention. They distinguish an investigable system from an apparatus of confirmation.
Belief does not become evidence by being automated
Belief has always shaped tools, institutions, architecture, and ritual. The new problem is not that technology has become contaminated by human meaning. Technology has never been free of human purposes.
The problem is that computation can hide interpretation behind precision.
Once a metaphysical assumption controls measurement, feedback, classification, access, or memory, it has entered the material conditions of another person’s experience. It may still be meaningful. It may be generative. It may become supported by evidence. It may also be wrong.
Its new power creates a new burden.
The belief must be named. Its status must be disclosed. Its effects must be observed. Alternatives must remain possible. Consent must cover the interpretation, not only the sensor. Data governance must protect inferred meaning, not only raw signals. Participants must retain authority over their own reports, and systems must not claim authority they have not earned.
Metaphysical Technology is being proposed as the category for designed technologies, environments, methods, and frameworks that deliberately engage questions, experiences, or relationships traditionally treated as metaphysical and bring them into operational, observable, testable, or governed form.
The final word matters here.
Governance is not what happens after belief is proved.
It is what becomes necessary when belief can act.
Sources
- Mattis Jacobs, Christian Kurtz, Judith Simon, and Tilo Böhmann, “Value Sensitive Design and Power in Socio-Technical Ecosystems,” Internet Policy Review 10, no. 3 (2021).
- Jay A. Olson et al., “Tripping on Nothing: Placebo Psychedelics and Contextual Factors,” Psychopharmacology 237 (2020): 1371–1382.
- National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, “The Belmont Report,” April 18, 1979.
- UNESCO, “Ethics of Neurotechnology.”
- Federal Trade Commission, “Bringing Dark Patterns to Light,” September 2022.
- National Institute of Standards and Technology, “Artificial Intelligence Risk Management Framework 1.0,” January 2023.