Cognitive Trajectory Modeling reframes co-creative human-AI interaction as trajectories across cognitive attractor landscapes, arguing that temporal patterns are only interpretable when states carry directional cognitive meaning; the proposal clarifies conceptual distinctions but remains to be operationalized and empirically validated.

Cognitive Trajectory Modeling: Quantifying Human-AI Co-Creation through Cognitively Grounded Interaction Trajectories

Nicholas Davis · June 13, 2026

arxiv theoretical n/a evidence 7/10 relevance Full text usable extracted full text Source PDF

Cognitive Trajectory Modeling (CTM) is a theoretical framework that models co-creative human-AI interaction as temporally organized cognitive trajectories across meaningful attractor landscapes and formalizes when temporal representations can be interpreted as cognitive trajectories.

Co-creative AI research increasingly seeks methods capable of representing how interaction dynamics evolve through time. While many existing approaches focus on observable interaction characteristics, interaction metrics, behavioral coding schemes, or activity traces, these methods often struggle to capture higher-order interaction dynamics, including how collaborative processes reorganize, stabilize, regulate, and evolve through time. This paper introduces Cognitive Trajectory Modeling (CTM) as a cognitive theory of interaction dynamics that conceptualizes cognition, interaction, and creative processes as temporally organized trajectories unfolding across cognitively meaningful attractor landscapes. CTM builds upon the theoretical foundations of the Enactive Model of Creativity and Creative Sense-Making (CSM), revisiting the role of sense-making curves and cognitive trajectories in representing co-creative interaction dynamics. We formalize this perspective through the Cognitive Trajectory Principle, which states that temporal representations are only theoretically interpretable as cognitive trajectories when their underlying states possess directional cognitive meaning. Building on this principle, CTM generalizes the notion of cognitive trajectories beyond any particular coding scheme and provides a broader framework for modeling interaction dynamics through trajectories unfolding across meaningful attractor landscapes. We further distinguish cognitive trajectories from interaction traces and situate CTM within a broader hierarchy of cognitive, interaction, and domain dynamics. More broadly, we argue that understanding co-creative systems requires methods capable of modeling how cognition and interaction dynamics unfold through time. CTM provides a foundation for studying interaction dynamics across co-creative AI and human-AI interaction.

Summary

Main Finding

Cognitive Trajectory Modeling (CTM) introduces a theory and methodological framework that represents human–AI co-creative interaction as cognitively meaningful trajectories unfolding across attractor landscapes. The core claim (the Cognitive Trajectory Principle) is that temporal interaction representations can only be interpreted as cognitive trajectories when their underlying states carry directional cognitive meaning — i.e., they map to distinct modes of sense-making rather than merely scalar or descriptive event counts. CTM formalizes this distinction, situates trajectories within a three-level hierarchy (cognitive, interaction, domain), and proposes trajectory-based tools for analyzing, comparing, and modeling co-creative dynamics.

Key Points

Cognitive Trajectory Principle: Temporal measures are only interpretable as cognitive trajectories if each state has directional cognitive significance (movement between qualitatively different modes of participation), not just magnitude.
Distinction: Interaction traces (time series of observable metrics/events) vs. cognitive trajectories (movement through a cognitively grounded state-space/attractor landscape).
Foundational participation coding (from Creative Sense-Making): three directional states — unclamped/exploratory (+1), clamped/exploitative (-1), and waiting/neutral (0). These are examples where accumulation yields interpretable trajectories.
Three-level analytical framework:
- Cognitive dynamics — organization of attention, perception, interpretation, sense-making.
- Interaction dynamics — coordination, regulation, co-participation among agents (including human–AI).
- Domain dynamics — evolution of artifacts, ideas, and domain-level outcomes.
CTM generalizes beyond any single coding scheme to multidimensional participation spaces, probabilistic trajectories, attractor landscapes, perceptual-logic trajectories, and adaptive regulation.
Historical and theoretical grounding: builds on enactive cognition, participatory sense-making, distributed cognition, improvisation studies, and prior work on Creative Sense-Making and quantified co-creation.
Methodological payoff: shifts analytical focus from final artifacts or isolated events toward temporal organization (stabilization, exploration, reorganization) of co-creative processes.

Data & Methods

Nature of contribution: primarily theoretical and methodological (framework development), not an empirical dataset study.
Conceptual tools:
- Sense-Making Curves: cumulative temporal representations designed to reflect movement across attractor landscapes.
- Directional vs. scalar state spaces: formal criteria for when time-series encodings represent directionally meaningful cognitive states.
- Attractor landscapes: structured state spaces with regions of stability, transition, and reorganization; trajectories map movement between these regions.
Example coding scheme (illustrative, from CSM):
- +1 Unclamped / exploratory participation
- -1 Clamped / committed/action-oriented participation
- 0 Waiting / neutral The numerical signs encode directionality (not just intensity).
Proposed extensions & methods:
- Multidimensional participation vectors (multiple cognitive axes)
- Probabilistic trajectory models (stochastic state transitions)
- Perceptual logic trajectories: mapping perception-affordance dynamics
- Adaptive regulatory dynamics: modeling how agents adapt their attractor landscapes over time
Analytical outputs: trajectory visualizations, comparative trajectory metrics, phase/attractor analysis, probabilistic/state-space inference; suitable for time-series, event-streams, and joint human–AI logs when states are cognitively meaningful.
Implementation note: CTM requires designing or validating coding/state representations that carry directional cognitive interpretation before interpreting accumulated temporal measures as trajectories.

Implications for AI Economics

Measurement of AI contribution: CTM provides process-level metrics that can separate AI and human roles across exploratory vs. exploitative phases, enabling finer-grained valuation of AI contributions to creative/innovative production beyond final outputs.
Productivity and complementarities:
- Trajectory-based measures allow economists to identify when AI augments human exploration (raising innovation potential) versus when it substitutes for human stabilization/execution.
- Can improve estimates of complementarity/substitutability by conditioning productivity on position in the cognitive/interaction state-space (e.g., human skills that matter at different trajectory phases).
R&D and innovation evaluation:
- Move from end-of-project outcome metrics to trajectory features (time spent exploring, frequency and depth of reorganizations, stability of attractors), which may better predict long-run innovation value.
- Useful for assessing interventions (tools, training, incentives) that change attractor landscapes or transition probabilities.
Labor-market effects and human capital:
- CTM helps identify which cognitive skills (managing exploration vs. execution, regulating co-creation) are complements to AI and therefore likely to command wage premia.
- Can inform upskilling strategies and job design by mapping tasks to cognitive trajectory roles.
Firm strategy, product design, and platforms:
- Firms can design AI tools to bias trajectories toward desirable regions (e.g., sustained exploration for early-stage innovation, or faster stabilization for production), affecting product development cycles, time-to-market, and costs.
- Platforms that expose interaction-trajectory analytics can create markets for “interaction intelligence” services (analytics, benchmarking).
Empirical research recommendations:
- Incorporate trajectory-state encodings into empirical models (difference-in-differences, event-study, structural models) to capture dynamic human–AI complementarities.
- Use randomized or quasi-experimental changes to AI tooling to observe shifts in state-space transitions and infer causal effects on innovation/productivity.
- Develop validation strategies to ensure coded states have directional cognitive meaning (survey, think-aloud, expert annotation).
Policy and welfare:
- Regulators and policy analysts should consider process-level impacts (changes in creative dynamics, shifts in required skills) when evaluating AI deployment, not only output-level metrics.
- CTM-informed metrics can be used to assess distributional effects across workers and firms (who benefits from altered attractor landscapes).
Market valuation and pricing:
- Pricing of co-creative AI services can incorporate trajectory value — e.g., premium for tools that robustly extend exploration or reliably speed productive stabilization.
- Investors and managers may value trajectory-improving features differently depending on stage (R&D vs. production).

Overall, CTM supplies a theoretically grounded way for AI economists to move beyond output-centric measures and to quantify the temporal, cognitive dimension of human–AI co-creation. That opens new paths for identifying complementarities, valuing AI contributions, designing contracts and training, and measuring the productivity and welfare effects of co-creative AI systems.

Assessment

Paper Typetheoretical Evidence Strengthn/a — The manuscript presents a conceptual/theoretical framework without empirical tests, causal identification, or quantitative evaluation, so no empirical strength can be assessed. Methods Rigormedium — The paper offers a formalized principle (the Cognitive Trajectory Principle) and situates the proposal within existing enactive theories of creativity, showing conceptual coherence and careful distinction of constructs, but it lacks operationalization, empirical validation, or robustness checks that would raise methodological rigor to high. SampleNo empirical sample or data — the work is a theoretical/conceptual development building on prior literature (Enactive Model of Creativity, Creative Sense-Making) and conceptual examples rather than applied datasets or experiments. Themeshuman_ai_collab innovation GeneralizabilityNo empirical validation — applicability to real-world human-AI interactions untested, Framework may be difficult to map to observable/measurable interaction traces across domains, Primarily oriented to co-creative contexts; may not generalize to other AI uses (e.g., automated decision systems), Relies on the assumptions of enactive theories of cognition, which may not be universally accepted, Practical implementation likely requires domain-specific coding schemes and design choices that could limit transferability

Claims (8)

Claim	Direction	Outcome	Confidence & Evidence	Details
Many existing approaches (interaction metrics, behavioral coding schemes, activity traces) often struggle to capture higher-order interaction dynamics, including how collaborative processes reorganize, stabilize, regulate, and evolve through time. Research Productivity	negative	ability to represent higher-order interaction dynamics	Reading fidelity high Study strength low	not reported 0.06
This paper introduces Cognitive Trajectory Modeling (CTM) as a cognitive theory of interaction dynamics that conceptualizes cognition, interaction, and creative processes as temporally organized trajectories unfolding across cognitively meaningful attractor landscapes. Research Productivity	positive	conceptual capacity to model temporally organized interaction dynamics	Reading fidelity high Study strength speculative	not reported 0.02
CTM builds upon the theoretical foundations of the Enactive Model of Creativity and Creative Sense-Making (CSM), revisiting the role of sense-making curves and cognitive trajectories in representing co-creative interaction dynamics. Research Productivity	positive	theoretical alignment with Enactive Model/CSM for representing co-creative dynamics	Reading fidelity high Study strength speculative	not reported 0.02
The Cognitive Trajectory Principle formalizes the perspective that temporal representations are only theoretically interpretable as cognitive trajectories when their underlying states possess directional cognitive meaning. Research Productivity	positive	theoretical interpretability of temporal representations as cognitive trajectories	Reading fidelity high Study strength speculative	not reported 0.02
CTM generalizes the notion of cognitive trajectories beyond any particular coding scheme and provides a broader framework for modeling interaction dynamics through trajectories unfolding across meaningful attractor landscapes. Research Productivity	positive	generality and applicability of interaction-dynamics modeling frameworks	Reading fidelity high Study strength speculative	not reported 0.02
The paper distinguishes cognitive trajectories from interaction traces and situates CTM within a broader hierarchy of cognitive, interaction, and domain dynamics. Research Productivity	positive	conceptual distinction between cognitive trajectories and interaction traces	Reading fidelity high Study strength speculative	not reported 0.02
Understanding co-creative systems requires methods capable of modeling how cognition and interaction dynamics unfold through time. Research Productivity	positive	necessity of temporal modeling methods for studying co-creative systems	Reading fidelity high Study strength low	not reported 0.06
CTM provides a foundation for studying interaction dynamics across co-creative AI and human-AI interaction. Research Productivity	positive	utility of CTM as foundational framework for co-creative AI and human-AI interaction research	Reading fidelity high Study strength speculative	not reported 0.02