Research topics

12 cross-cutting research topics

Central questions of the twelve topics

1. 01
   Emergent Intelligence

   When does interaction stop being merely interaction and begin producing persistent intelligent structure?

2. 02
   Human–AI Co-Evolution

   How does repeated interaction change the human's use of the system and the structure of the interaction itself?working formulation

3. 03
   Cognitive Systems

   How can cognitive states be represented computationally?working formulation

4. 04
   Neuroadaptive Human–AI Systems

   How can systems adapt to changing cognitive states while keeping measured and inferred states apart?working formulation

5. 05
   Multi-Agent and Collective Intelligence

   When does orchestration produce collective capabilities?working formulation

6. 06
   Memory, State and Identity

   What should a system keep over time, and when does memory produce false continuity?working formulation

7. 07
   Knowledge, Evidence and Epistemic Integrity

   How does observed information stay distinguishable from inferred information when a system accumulates knowledge over a long time?working formulation

8. 08
   World Models

   How does a world model stay traceable while it changes with new observations?working formulation

9. 09
   Interaction Dynamics

   How do interaction patterns evolve across extended periods, beyond individual conversations?working formulation

10. 10
    Emergence Detection

    What shows that a behavioral change is emergence?working formulation

11. 11
    Adaptive Systems

    How does adaptive behavior stay inspectable and traceable?working formulation

12. 12
    Human–Machine Interfaces

    How do interfaces shape cognition, trust and system understanding, and what will an interface need to show?working formulation

The twelve topics

1. 01

   Emergent Intelligence

   • Main space: C Emergence, Adaptation & Complex Dynamics

   Open: Emergent Intelligence

   Emergent Intelligence

   ×

   When does interaction stop being merely interaction and begin producing persistent intelligent structure?

   Investigates how intelligent behavior can arise from interactions between simpler processes.

   emergence · adaptive behavior · self-organization · recursive feedback · distributed cognition · collective intelligence · strategy formation · system-level capabilities · behavioral transitions · stability and instability of intelligent behavior

   Spaces touched

   • E Distributed, Agentic & Collective Intelligence
   • A Human–AI Interaction & Co-Evolution

   Programs

   • Emergence & Stabilization
   • Distributed Intelligence
2. 02

   Human–AI Co-Evolution

   • Main space: A Human–AI Interaction & Co-Evolution

   Open: Human–AI Co-Evolution

   Human–AI Co-Evolution

   ×

   How does repeated interaction change the human's use of the system and the structure of the interaction itself?working formulation

   Human–AI interaction was one of the original experimental foundations of EIL and remains a major topic of the research. EIL investigates how repeated interaction changes both the human's use of the system and the structure of the interaction itself. Human and AI form a changing coupled system.

   development of interaction strategies · changes in prompt behavior · cognitive offloading · externalized reasoning · iterative problem decomposition · reflective AI use · human adaptation to model behavior · model-context adaptation · interaction routines · semantic compression · shared working representations · emergent collaboration structures

   Programs

   • Human–AI Co-Evolution
   • Longitudinal Interaction Impact Evaluation
3. 03

   Cognitive Systems

   • Main space: B Human Cognition, Behavior & Externalization

   Open: Cognitive Systems

   Cognitive Systems

   ×

   How can cognitive states be represented computationally?working formulation

   A major direction of EIL concerns the relationship between artificial systems and human cognition. It connects Human–AI interaction with cognitive science and adaptive system design.

   representation of cognitive states · reasoning strategies · problem-solving behavior · cognitive load · uncertainty · externalized reasoning

   Further questions of the topic (5)

   • How do reasoning strategies change over time?
   • How can longitudinal interaction reveal changes in problem-solving behavior?
   • Which interaction patterns correlate with cognitive load or uncertainty?
   • How does externalized reasoning modify internal reasoning?
   • How can adaptive systems react to cognitive state without over-interpreting the user?

   Spaces touched

   • F Computational Interaction & Cognitive Modeling

   Programs

   • Emergent Cognitive Architecture
   • Cognitive Externalization
4. 04

   Neuroadaptive Human–AI Systems

   • Main space: G Human–Machine Interfaces & Neuroadaptive Systems

   Open: Neuroadaptive Human–AI Systems

   Neuroadaptive Human–AI Systems

   ×

   How can systems adapt to changing cognitive states while keeping measured and inferred states apart?working formulation

   A developing direction extends the EIL approach toward neuroadaptive systems. Future systems may take physiological or neurophysiological signals into account in addition to explicit commands and observable interaction patterns. The long-term goal is systems that adapt to changing human cognitive states while keeping strict epistemic boundaries between measured, observed, inferred and unknown states.

   EEG · attention-related signals · cognitive workload indicators · physiological state · interaction latency · behavioral dynamics · task performance · uncertainty patterns

   Bridge: Human–AI interaction → cognitive state modeling → neuroadaptive Human–AI systems.

   Spaces touched

   • B Human Cognition, Behavior & Externalization
   • A Human–AI Interaction & Co-Evolution

   Programs

   • Neuroadaptive Human–AI Systems
5. 05

   Multi-Agent and Collective Intelligence

   • Main space: E Distributed, Agentic & Collective Intelligence

   Open: Multi-Agent and Collective Intelligence

   Multi-Agent and Collective Intelligence

   ×

   When does orchestration produce collective capabilities?working formulation

   EIL investigates intelligence distributed across multiple specialized agents. The deeper questions concern coordination. Multi-agent systems function as experimental environments for studying distributed intelligence.

   specialization · distribution of responsibility · state transfer · conflict resolution · provenance · orchestration · semantic drift · collective intelligence

   Further questions of the topic (7)

   • How should specialization emerge?
   • How are responsibilities distributed?
   • How do agents transfer state?
   • How are conflicting conclusions resolved?
   • How can provenance survive multi-agent reasoning?
   • How does local agent behavior affect global system behavior?
   • How can agent systems avoid uncontrolled semantic drift?

   Programs

   • Distributed Intelligence
6. 06

   Memory, State and Identity

   • Main space: E Distributed, Agentic & Collective Intelligence

   Research theme, no program of its own yet

   Open: Memory, State and Identity

   Memory, State and Identity

   ×

   What should a system keep over time, and when does memory produce false continuity?working formulation

   Persistent intelligence needs more than a sequence of isolated responses. EIL investigates structures that let systems keep continuity over time: working memory, long-term memory, episodic memory, semantic memory, system state, agent state, interaction history, identity continuity, task state and environmental state. Persistent memory can raise capability and also create risks. Memory is treated as an active system component.

   memory types · agent state · identity continuity · what should persist · outdated beliefs · false continuity · feedback amplification · hallucinated causality · state contamination · identity drift

   Spaces touched

   • H Measurement, Evidence & Research Infrastructure
   • A Human–AI Interaction & Co-Evolution
7. 07

   Knowledge, Evidence and Epistemic Integrity

   • Main space: H Measurement, Evidence & Research Infrastructure

   Research theme, no program of its own yet

   Open: Knowledge, Evidence and Epistemic Integrity

   Knowledge, Evidence and Epistemic Integrity

   ×

   How does observed information stay distinguishable from inferred information when a system accumulates knowledge over a long time?working formulation

   An intelligent system needs to know what kind of information it has. EIL investigates epistemic structures that distinguish observed, inferred, hypothesis, simulated, derived, unknown and contradicted. Without epistemic separation, inference gradually becomes indistinguishable from evidence. EIL treats provenance, evidence chains, source attribution, uncertainty, contradiction detection, confidence and temporal validity as fundamental components of intelligent systems.

   epistemic labels · provenance · evidence chains · source attribution · uncertainty · contradiction detection · confidence · temporal validity

   Spaces touched

   • D Reconstruction, Perspective & World Models
8. 08

   World Models

   • Main space: D Reconstruction, Perspective & World Models

   Open: World Models

   World Models

   ×

   How does a world model stay traceable while it changes with new observations?working formulation

   Intelligent action requires some representation of the world. EIL investigates how systems construct and maintain world models from incomplete and changing observations. A world model may contain entities, relations, states, events, causal hypotheses, temporal dependencies, unresolved relationships, predictions and uncertainty. The research challenge is a representation that can change and stay traceable.

   state reconstruction · temporal reasoning · causal modeling · graph structures · prediction · unresolved dependencies · anomaly detection · model revision

   Programs

   • World / Actor / Causal Model Reconstruction
   • Perspective Modeling & Cognitive Reconstruction
9. 09

   Interaction Dynamics

   • Main space: A Human–AI Interaction & Co-Evolution

   Open: Interaction Dynamics

   Interaction Dynamics

   ×

   How do interaction patterns evolve across extended periods, beyond individual conversations?working formulation

   Interaction itself is treated as a measurable system. Longitudinal interaction data shows how patterns evolve across extended periods, beyond individual conversations.

   turn structure · information density · semantic shifts · behavioral transitions · tool use · delegation · correction cycles · feedback loops · interaction depth · role formation · task decomposition

   Spaces touched

   • C Emergence, Adaptation & Complex Dynamics
   • F Computational Interaction & Cognitive Modeling

   Programs

   • Human–AI Co-Evolution
   • Longitudinal Interaction Impact Evaluation
10. 10

    Emergence Detection

    • Main space: C Emergence, Adaptation & Complex Dynamics

    Open: Emergence Detection

    Emergence Detection

    ×

    What shows that a behavioral change is emergence?working formulation

    Whether a behavioral change is emergence is decided by longitudinal evidence. EIL investigates methods for identifying potentially emergent phenomena.

    recurring strategies · novel coordination structures · unexpected capability combinations · persistent behavioral patterns · new interaction protocols · self-stabilizing workflows · semantic conventions · role differentiation

    Observation first. Interpretation second. Claim last.

    Spaces touched

    • F Computational Interaction & Cognitive Modeling

    Programs

    • Emergence & Stabilization
11. 11

    Adaptive Systems

    • Main space: C Emergence, Adaptation & Complex Dynamics

    Research theme, no program of its own yet

    Open: Adaptive Systems

    Adaptive Systems

    ×

    How does adaptive behavior stay inspectable and traceable?working formulation

    EIL studies systems whose behavior changes according to user behavior, interaction history, system state, environmental state, confidence, uncertainty, task context and cognitive state. Adaptation can occur at seven levels. The research goal is adaptive behavior that remains inspectable and traceable.

    interface adaptation · interaction adaptation · workflow adaptation · agent adaptation · memory adaptation · model-selection adaptation · cognitive-state adaptation

    Spaces touched

    • G Human–Machine Interfaces & Neuroadaptive Systems
    • A Human–AI Interaction & Co-Evolution
12. 12

    Human–Machine Interfaces

    • Main space: G Human–Machine Interfaces & Neuroadaptive Systems

    Research theme, no program of its own yet

    Open: Human–Machine Interfaces

    Human–Machine Interfaces

    ×

    How do interfaces shape cognition, trust and system understanding, and what will an interface need to show?working formulation

    EIL keeps a strong Human–Machine Interaction foundation. Interfaces shape cognition, decision-making, attention, error propagation, trust, system understanding and interaction strategies. That links classic HMI/UX research directly to emerging intelligent systems. Future intelligent interfaces may need to show system state, agent activity, evidence, uncertainty, memory, provenance, reasoning structures and conflicting interpretations, next to controls and information.

    cognition · decision-making · attention · error propagation · trust · system understanding · interaction strategies · display of system state, evidence, uncertainty and provenance

    Spaces touched

    • A Human–AI Interaction & Co-Evolution