Enactive AI: Adaptive Intelligence Under Structural Drift

My research investigates how intelligent systems maintain coherent behavior in environments characterized by structural change. Real-world environments rarely remain stationary: correlations shift, signals decay, and contextual structures reorganize over time. Traditional AI systems often assume stable conditions and therefore struggle when underlying dynamics evolve.

I develop enactive AI architectures that regulate perception, sense-making, and action through continuous interaction with changing environments. These systems do not rely solely on fixed training assumptions or static optimization objectives. Instead, they maintain coherence by detecting structural change and reorganizing their engagement with the environment accordingly.

This research program explores how adaptive intelligent systems can operate effectively under non-stationarity across multiple domains, including financial systems, human–AI collaboration, organizational decision environments, and distributed socio-technical systems.

The program is organized around four complementary research thrusts.

1. Structural Drift Detection

The first thrust investigates how intelligent systems can detect changes in the underlying structure of their environment.

Many real-world systems operate on streaming data where statistical relationships evolve over time. Traditional machine learning methods often assume stationary distributions, leading to degraded performance when environmental structure shifts.

This research develops methods for identifying structural drift in dynamic data streams, enabling systems to recognize when previously learned assumptions no longer hold. By treating drift as an informational signal rather than noise, adaptive systems can detect emerging regimes, reorganizing their interpretation of the environment and preparing for new behavioral responses.

2. Enactive AI Architectures

The second thrust develops computational architectures inspired by the enactive paradigm in cognitive science.

In this view, intelligent behavior emerges through continuous interaction between an agent and its environment. Effective systems therefore maintain coherence not through static prediction alone, but through feedback-regulated loops connecting perception, interpretation, and action.

This work explores architectural principles for building adaptive AI systems that regulate their engagement with dynamic environments. These architectures support ongoing sense-making processes that allow systems to interpret environmental change and reorganize their behavioral strategies accordingly.

3. Human–AI Participatory Sense-Making

The third thrust investigates how humans and intelligent systems coordinate their behavior in shared environments.

Human cognition itself operates through processes of participatory sense-making, where meaning emerges through interaction between agents. Understanding how artificial systems can participate in these processes is critical for designing AI technologies that collaborate effectively with people.

This research studies interaction dynamics in human–AI systems, developing models that quantify coordination patterns, adaptive turn-taking, and collaborative sense-making processes. These insights inform the design of AI systems that support human decision-making rather than replace it.

4. Adaptive Systems in Real-World Environments

The fourth thrust investigates how adaptive architectures operate within real-world domains characterized by structural change.

To explore these questions empirically, I design experimental computational platforms that function as research testbeds for adaptive intelligence. One such system, the Emergence Machine, examines how adaptive decision architectures can detect structural drift in streaming environments and reorganize perception, interpretation, and action in response.

These platforms allow theoretical models from cognitive science to be evaluated in operational contexts, including financial markets, organizational decision systems, and distributed human–AI environments. By grounding theoretical principles in working systems, this research aims to develop adaptive AI architectures capable of sustaining coherent behavior across a wide range of dynamic environments.

This program advances a broader goal: the development of intelligent systems capable of maintaining coherence under structural change, bridging insights from cognitive science, artificial intelligence, and complex adaptive systems research.

Research Focus

My work explores how intelligent systems maintain coherent behavior in dynamic environments. I design interactive AI architectures grounded in cognitive theory, particularly the enactive perspective that cognition emerges through continuous interaction between agent and environment.

Across my research, I develop computational systems that:

I operate at the intersection of cognitive science, AI system design, human–computer interaction, and adaptive decision systems, translating theoretical models of sense-making into implementable computational architectures.