Cognitive AI Research
Beyond automation.
Towards cognition.
A separate pillar from consulting. Research at the frontier of what AI systems can do: memory, reasoning, consciousness, emergence.
The Research Question
Why current AI systems hit a ceiling.
Today's AI systems are extraordinary retrieval machines with a generation layer. They are not reasoning systems. They have no persistent memory. They cannot hold structured understanding of relationships between ideas.
This research track explores the architectural foundations for systems that can genuinely reason — drawing from cognitive science, information theory and systems architecture.
Systems over models.
Architecture over prompting.
Reasoning over retrieval.
Memory over context windows.
Focus Areas
What the research covers.
Cognitive Architectures
Exploring separation of memory, reasoning and output as distinct architectural concerns, showing how an AI system's structure can mirror layered cognition.
Memory Systems
What does persistent, structured, contextually-weighted memory look like at the architectural level, and how does it transform reasoning quality?
Reasoning Frameworks
Designing explicit reasoning layers that operate on structured knowledge rather than implicitly within a generation pass.
Human-AI Interaction
Interaction models for systems with genuine reasoning and memory, where the human and the system co-construct understanding.
Consciousness-Inspired Computing
Applying attention, integration and predictive processing from consciousness research as design inspiration for AI architectures.
Emergence & Knowledge Representation
How complex behaviours emerge from structured knowledge representations, at the boundary between retrieval and genuine understanding.
Active Projects
Where the research lives.
Numocore
Consciousness-based AI pipeline
An AI pipeline built around consciousness-inspired architecture, a chatbot system that reasons from awareness primitives rather than pure retrieval. More details to follow.
Massive Dynamics
AI company · CTO & Co-founder
Co-founded AI-native company building at the intersection of advanced systems and applied intelligence. Research and product intersect here.
Sentients Research
The studio's research track
The broader research practice behind Sentients, connecting cognitive science, information theory and systems architecture to build foundations for AI systems beyond current paradigms.
Interested in collaborating
on this research?
Deep-tech collaborations, research partnerships and university engagements welcome.
Get in touch