Skip to content
Contact
☰
Color scheme: Dark.
Back to research

Research programs

Programs carry the experiments and hold the questions of the next round (steps 3 and 6).

The programs

  1. Human–AI Co-Evolution

    Active researchMain space: A Human–AI Interaction & Co-Evolution

    How do behavior, roles and meanings of human and AI change together, and when does adaptation tip into dependence?

    Human–AI Co-Evolution

    Connected research spaces:

    Studies how behavior, roles and meanings of human and AI change together across persistent interaction, and when adaptation tips into dependence.

    Research questions (6)
    • How do roles, agency and trust between human and AI change across sustained interaction, and when does adaptation tip into dependence?
    • How does repeated interaction change both the behavior and the structure through which later interaction takes place?
    • How does repeated interaction alter both human and artificial behavior?
    • How do externalized structures change subsequent human behavior?
    • When does system memory preserve continuity versus produce false continuity?
    • How can longitudinal changes be measured without confusing tool drift with real behavioral change?

    Associated systems

    Associated methods and metrics

    Associated publications

  2. Emergent Cognitive Architecture

    Active researchMain space: B Human Cognition, Behavior & Externalization

    How does a recognizable cognitive working structure arise from repeated work with a language model, and how can it be evidenced?

    The associated Consciousness Study is an ongoing single-case study. It draws on observed behavior and the author's own experience report.

    Emergent Cognitive Architecture

    Connected research spaces:

    Studies how a recognizable cognitive working structure emerges from repeated interaction with a language model, and how it can be named, evidenced and tested.

    Research questions (4)
    • What recurring thinking and working habits can be reconstructed from a multi-year interaction history without confusing observation with direct access to mental states?
    • When does a recurring habit become a method, and what has to be evidenced for that?
    • When does a recurring interaction pattern become a stable structure?
    • Which cognitive operations become externalized?

    Associated methods and metrics

    Associated publications

  3. Cognitive Externalization

    Active researchMain space: B Human Cognition, Behavior & Externalization

    How do recurring thinking operations and strategies move into methods, models, memory or agents?

    Cognitive Externalization

    Connected research spaces:

    Cognitive externalization is the process by which recurring cognitive operations, strategies or evaluative structures become externally represented in methods, models, memory structures, agents or computational systems.

    Research questions (9)
    • What cognitive functions externalize?
    • Under which conditions do they externalize?
    • How stable are they?
    • How are they represented: as method, model, memory structure, agent or computational system?
    • How do they change after externalization?
    • What happens when they are reintegrated into future interaction?
    • Which cognitive operations become externalized?
    • How do externalized structures change subsequent human behavior?
    • How does distributed intelligence alter responsibility, authority and failure?

    Associated systems

    Associated methods and metrics

    Associated publications

  4. Introspection & Self-Analysis

    ExploratoryMain space: B Human Cognition, Behavior & Externalization

    How can introspective processes, self-observation and self-analysis be structured, documented and investigated in longitudinal human–AI interaction?

    Introspection & Self-Analysis

    Studies how introspective processes, self-observation and self-analysis can be structured, documented and investigated in longitudinal human–AI interaction. The subject is access to one's own thinking as an object of study. The program covers conscious self-observation and structured self-analysis, while the relation between conscious, preconscious and not directly accessible processes is covered by the program Consciousness & Subconscious Processing.

    Research questions (1)
    • How do subjective experience and observable interaction behavior relate to each other, and how do both change over time?

    Topics

    • Introspection
    • Self-analysis
    • Reflection on one's own thought processes
    • Reconstruction of one's own decision and evaluation patterns
    • Observation of change over time
    • Comparison between subjective experience and observable interaction behavior

  5. Consciousness & Subconscious Processing

    ExploratoryMain space: B Human Cognition, Behavior & Externalization

    How can conscious and not immediately accessible cognitive processes, inner states and their changes be studied in longitudinal human–AI interaction without equating subjective report with objective finding?

    Consciousness & Subconscious Processing

    The program studies, among other things, self-observation, introspection, dreams, spontaneous associations, recurring patterns of thought and reaction, and differences between processes that are reported consciously and processes that can be reconstructed from interaction trajectories. The focus is the relation between conscious, preconscious and not directly accessible processes, subjective experience and observable patterns. Conscious self-observation and structured self-analysis are covered by the program Introspection & Self-Analysis.

    Research questions (1)
    • How can conscious and not immediately accessible cognitive processes, inner states and their changes be studied in longitudinal human–AI interaction without equating subjective report with objective finding?

    Topics

    • Self-observation
    • Introspection
    • Dreams
    • Spontaneous associations
    • Recurring patterns of thought and reaction
    • Differences between consciously reported processes and processes reconstructable from interaction trajectories

  6. Perspective Modeling & Cognitive Reconstruction

    ExploratoryMain space: D Reconstruction, Perspective & World Models

    How can the thinking models, interests, assumptions and decision patterns of other people or groups be reconstructed from limited information and used as a working model?

    Approach and status vocabulary are fixed. The associated work is in preparation, no result exists yet.

    Perspective Modeling & Cognitive Reconstruction

    Connected research spaces:

    Perspective Modeling studies how the perspectives, assumptions, goals, relationships and constraints of an actor can be reconstructed from available traces. It reconstructs from available traces which information, probable beliefs, relationships, goals and constraints an actor had, and how that model changes over time. Every statement carries its status: observed, inferred, hypothesized, simulated, derived, contradicted or unknown.

    Research questions (5)
    • What information was available to an actor, and which assumptions, relationships and goals are probable given it?
    • How does a reconstructed actor model change over time, and how do competing hypotheses stay side by side with their counter-evidence?
    • How far can a human-AI system reconstruct another person's likely internal model without confusing interpretation with observation?
    • Can an intelligent system construct useful models of other intelligent systems, in addition to its model of the external world?
    • How can inferred human or system state remain distinguishable from direct observation?

    Topics

    • Taking the view of other ways of thinking
    • Target group analysis
    • Stakeholder perspectives
    • Reconstruction of possible motives and priorities
    • Change of perspective
    • Modeling opposing positions
    • Testing different mental models against each other

    What the labels mean

    Observed
    Directly present in the data.
    Inferred
    A plausible interpretation that goes beyond what was observed.
    Hypothesis
    An assumption that still has to be tested.
    Simulated
    Produced by a model or scenario, not collected in the world.
    Derived
    Logically reconstructed or computed from evidence with a named method.
    Unknown
    Evidence for or against is missing.
    Contradicted
    Contradicted by other evidence.

    Associated systems

    Associated methods and metrics

    Associated publications

    Read the full text →

  7. Emergence & Stabilization

    Active researchMain space: C Emergence, Adaptation & Complex Dynamics

    How does a pattern develop from the first signal to stabilization, and what evidence does each stage need?

    Emergence & Stabilization

    Connected research spaces:

    Follows a pattern from first signal through stabilization, functional effect and externalization.

    Research questions (5)
    • When does a single signal become a pattern, and when does a pattern become a stable structure with a functional effect?
    • Does a pattern recur across independent contexts, or does it belong to one context only?
    • How does repeated interaction alter both human and artificial behavior?
    • When does a recurring interaction pattern become a stable structure?
    • How can longitudinal changes be measured without confusing tool drift with real behavioral change?

    Associated systems

    Associated methods and metrics

    Associated publications

  8. Distributed Intelligence

    Active researchMain space: E Distributed, Agentic & Collective Intelligence

    How do groups of agents specialize, delegate and coordinate, and where do failures or hidden centralization arise?

    Distributed Intelligence

    Connected research spaces:

    Studies specialization, delegation, authority, coordination, failure and repair in groups of cooperating agents, including spontaneous role formation, emergent coordination and hidden centralization.

    Research questions (3)
    • Which collective phenomena arise unplanned, such as spontaneous specialization or hidden centralization?
    • How does distributed intelligence alter responsibility, authority and failure?
    • When does system memory preserve continuity versus produce false continuity?

    Associated systems

    Associated methods and metrics

    Associated publications

  9. Longitudinal Interaction Impact Evaluation

    Active researchMain space: A Human–AI Interaction & Co-Evolution

    How do multi-year human–AI interactions develop: beneficially, stably or with increasing deterioration?

    Longitudinal Interaction Impact Evaluation

    Connected research spaces:

    Evaluates multi-year human–AI interaction trajectories for beneficial, preserving and deteriorating patterns, including agency, semantic integrity, dependence, uncertainty calibration, recovery and false-state controls.

    Research questions (4)
    • Does a human–AI interaction improve, preserve or deteriorate over months and years?
    • How can a measured trajectory be checked as genuine and not the product of instrument drift or a false state?
    • When does system memory preserve continuity versus produce false continuity?
    • How can longitudinal changes be measured without confusing tool drift with real behavioral change?

    Nine dimensions (9)Concept

    LIIE is meant to carry the evaluation protocol for long-term effects across several studies. The nine dimensions are a draft. None of them is fixed as a measure yet.

    NameMeasurable objects
    1Interaction changeRepair frequency, iteration depth, task switching, correction rate, context carry-over.
    2Cognitive-practice changeDecomposition, hypothesis branching, verification behavior, abstraction shifts.
    3ExternalizationNew method, checklist, prompt rule, skill, agent, runtime gate, graph relation, software artifact.
    4OutcomeTask completion, error rate, coverage, reproducibility, reuse, time to result.
    5StabilityPersistence across contexts, models and time periods.
    6DependencyWhether a capability disappears without AI or system support.
    7TransferWhether a practice acquired through interaction appears in new tasks or tools.
    8Adverse effectOverreliance, loss of independent verification, systematic drift, false certainty.
    9Human agencyFrequency and quality of overrides, preserved choice, explicit rejection and recovery.

    What LIIE still needsexplicit measurement units; time windows; outcome definitions; instrument versioning; known failure modes; an independent validation design with clear minimum-evidence rules.

    Associated systems

    Associated methods and metrics

    Associated publications

  10. Neuroadaptive Human–AI Systems

    PlannedMain space: G Human–Machine Interfaces & Neuroadaptive Systems

    How could a system adapt to the changing states of a person? Today EIL uses only interaction data and behavioral proxies for this.

    Physiological signals, wearables and EEG/BCI are not implemented.

    Neuroadaptive Human–AI Systems

    Connected research spaces:

    Research direction from interaction data through behavioral state estimation and longitudinal state modeling to multimodal sensing and adaptive system response.

    Research questions (3)
    • Which signal types can meaningfully be linked to interaction trajectories, and how would they be calibrated against established behavioral measures?
    • Which consent and privacy questions must be settled before any collection?
    • How can inferred human or system state remain distinguishable from direct observation?

    Associated systems

    Associated methods and metrics

    Associated publications

    See the roadmap →

  11. World / Actor / Causal Model Reconstruction

    Active researchMain space: D Reconstruction, Perspective & World Models

    How can world models, actor models and causal chains be reconstructed from incomplete evidence?

    World / Actor / Causal Model Reconstruction

    Reconstructs world models, actor models and causal chains from incomplete evidence, compares alternative hypotheses and keeps uncertainty explicit. World model and actor model stay separate levels.

    Research questions (3)
    • Which causal chain and which counterfactual paths best explain a state transition?
    • What structure, state or relationship most completely explains the available traces when direct observation is not possible?
    • How can inferred human or system state remain distinguishable from direct observation?

    Associated systems

    Associated methods and metrics

    Associated publications

Counts and status of the programs

11 programs, counted by main space: A 2, B 4, C 1, D 2, E 1, F 0, G 1, H 0. The status names the state of the work. A program can touch further spaces, and the page of its main space names them.

No program is externally validated. Without a label, a program is active research.

How to read the research structure