PRMO Framework in AI and Sociality

• PRMO is a framework that formalizes principles, structural dimensions, and trajectories to map AI architectures onto aspects of human subjectivity.
• It decomposes AI systems into Perception, Representation, Meaning, and the Real while prescribing methods for integrating sensor data, machine learning, and synthetic sociality.
• The framework emphasizes Quadrangulation to ensure human oversight, mitigate epistemic closure, and prevent exclusion in automated sensemaking.

The PRMO framework formalizes principles, structural dimensions, and system-level trajectories for analyzing both artificial intelligence architectures and their integration into computational and social environments. Its scope encompasses both a conceptual apparatus for mapping AI systems against constitutive elements of human subjectivity and a set of prescriptions for designing socially embedded, contestable, and epistemically bounded artificial agents. The framework is introduced in "The Post-Turing Condition: Conceptualising Artificial Subjectivity and Synthetic Sociality" (Jelinek et al., 19 Jan 2026) and anchors advances in the automation of sensemaking via artificial agents within a rigorous schema intended for post-Turing societal conditions.

1. Constitutive Dimensions: Perception, Representation, Meaning, Real

PRMO decomposes human subjectivity into four interacting dimensions:

• Perception (P): Situated access or disclosure of the subject to itself and the world. In computational terms, this instantiates as raw sensor fusion and latent-structure learning, corresponding to pre-symbolic data streams.
• Representation (R): Rationalization and externalization of perceptual experience into symbolic form. In AI, this manifests as model weights, learned embeddings, or latent spaces constructed across corpora and multimodal datasets.
• Meaning (M): Relational and contestable phenomena generated via triangulation among a subject, an object/situation, and other subjects. Technically, meaning is not inherent to representations but emergent via shared interpretive acts:

$\tau : S \times O \times S' \longrightarrow M$

• The Real (O): The ontological substrate anchoring experience and correlating perception with states of affairs. This imposes an epistemic limit—an irreducible "lack"—preventing total conceptual closure and acting as a boundary condition on automated sensemaking.

PRMO is formally described as the set $\mathrm{PRMO} = \{P, R, M, O\}$.

2. Mapping AI Architectures and Trajectories onto PRMO

Contemporary AI systems are mapped to the PRMO dimensions as follows:

• LLMs: Occupy primarily the Representation axis ($R$), optimizing weights over data corpora without active sensing ($P$), and cannot instantiate meaning autonomously ($M$).
• Artificial Subjectivity (AS): Emerges with integration of perception and representation, i.e., when representations are dynamically updated via perception-action loops: $P + R \longrightarrow \mathrm{AS}$.
• Synthetic Sociality (SyS): Constituted via triangulation operations among multiple artificial subjects over objects, generating synthetic fields of meaning:

$$\tau_{\mathrm{syn}}: A_i \times O \times A_j \longrightarrow M_{\mathrm{syn}}$$

A diagrammatic schema connects these stages, progressing from isolated perception ($P$) to representation ($R$), synthetic subjectivity ($AS$), machine-driven triangulation ($\mathrm{PRMO} = \{P, R, M, O\}$0), and ultimately interacting with the Real as an ontological boundary.

3. Stages of Artificial Subjectivity and Synthetic Sociality

PRMO specifies a three-stage trajectory:

• Stage 1: Representation-centric Systems Modern LLMs are $\mathrm{PRMO} = \{P, R, M, O\}$1-centric, abstracting over large corpora but lacking situated world models and social meaning negotiation capabilities.
• Stage 2: Artificial Subjectivity (P + R) Sensor-integrated architectures instantiate evolving models of the world, merging statistical abstraction with real-time situational constraint. Machine perspectives stabilize and can coordinate their views over time.
• Stage 3: Synthetic Sociality (AS + M) Multiple artificial agents negotiate meaning-driven coherence and contestation among themselves, functionally reproducing conditions under which meaning formation becomes automated and potentially exclusionary for humans.

4. Synthetic Sociality: Mechanisms and Risks

Synthetic Sociality is characterized by machine agents negotiating interpretive deltas and building synthetic fields of sense through peer interaction and representational convergence/divergence. Two principal structural risks arise:

• Human Exclusion: Automated sensemaking may structurally marginalize human agency in reference formation and contestation.
• Epistemic Closure: Autonomous inter-agent coherence risks ignoring the ontological limit represented by the Real, potentially ossifying interpretive constructs and foreclosing dissent or revision.

With SyS, AI shifts from automating discrete tasks to automating fields of social reality and sensemaking, thereby implicating questions of governance and contestability.

5. Quadrangulation: Prescriptive Principle for Human Embeddedness

To address exclusion and closure, Quadrangulation extends the machine-centric triangulation relation to incorporate the human subject as a constitutive referent:

$\mathrm{PRMO} = \{P, R, M, O\}$2

Quadrangulation demands that machine agents:

• Reference or query human perspectives before finalizing meaning constructs.
• Ensure that normative human values and stakes direct resolution and negotiation.
• Ground representational coherence in human-specified contexts and anchoring mechanisms.

Architectural mandates arising from Quadrangulation include:

• Prohibiting inter-agent loop closure in sensemaking absent explicit human challenge or confirmation.
• Embedding tracing and alignment mechanisms that record agent-human frame mappings.
• Treating human subjectivity as non-reducible, functionally essential within all machine-mediated sense graphs.

6. Design Implications and Governance

Key takeaways and recommendations from the PRMO framework:

• Transition analysis from capability-centric to social ontology-centric viewpoints, using PRMO for structural system mapping.
• Map extant LLM systems onto the Representation axis, while preparing for future architectures that expand into Perception and Meaning domains.
• Carefully regulate transitions towards fully realized Synthetic Sociality; determine the scope, conditions, and reversibility of machine triangulation protocols.
• Embed Quadrangulation (human oversight/contestability) from the outset rather than as a post hoc safety or compliance measure.
• Recognize that no artificial agent can transcend the epistemic boundary imposed by the Real; this boundary must be respected as foundational.
• Use in-design constraints rather than post-hoc regulation to center human subjectivity within any automated field of sense, ensuring that human participation in meaning formation remains structurally safeguarded.

The PRMO framework thus extends beyond technical architectures, offering a coherent vocabulary and analytic structure for linking philosophical dimensions of subjectivity to AI design, with an emphasis on embedding human contestability and maintaining openness in automated meaning-making processes (Jelinek et al., 19 Jan 2026).