Current program · 2026
Below outputs. Into structure.
Research on cognitive integrity, artificial-agent assessment, governability, complex systems, and consciousness—moving between foundational theory and deployable measurement.
View Google Scholar
Neural Axis Research
Research into the architecture of intelligence.
The conditions that keep human and artificial intelligence governable, reality-linked, and meaningfully self-directing.
Current program · 2026
The current research program.
Six connected lines of work examine intelligence below surface outputs: how cognitive boundaries hold, how agentic systems become governable, how trajectories become risky, and how complex forms acquire structure over time.
01Preprint · 2026
The First Infrastructure of Intelligence: Cognitive Integrity in Human–AGI Systems
Introduces cognitive integrity as the capacity to maintain calibrated attention, trust, contestability, and decision under pressure as human and institutional cognition couples to AI.
Read source02Collaborative research · 2026
The Case for a Neuropsychology of Artificial Agents
Proposes construct-valid assessment of cognitive, affective, personality, response-validity, and predictive architecture in artificial agents.
Read source03Preprint · 2026
Path-Sensitive AGI Alignment
Reframes alignment as a path-dependent problem shaped by trajectories, transitions, escape costs, and the capacity to remain governable.
Read source04Preprint · 2026
Structural Governability Beyond Output Alignment
Argues that aligned outputs are insufficient when an agent's internal and operational structure may resist oversight, correction, or safe reintegration.
Read source05Preprint · 2026
Morphology, Seam Topology, and Temporal Scaffolding
Examines how complex forms persist and change through morphology, boundaries, seams, and temporal scaffolding.
Read source06Preprint · 2026
Structural Preconditions of Basin Formation
Investigates the structural conditions under which stable basins, attractors, and organized patterns can form.
Read sourceAgentic neuropsychology
From benchmark scores to psychological architecture.
Artificial agents increasingly operate in medicine, law, finance, education, and government. Yet most evaluation still measures outputs, capabilities, or isolated failure modes. Agentic neuropsychology asks what stable or context-sensitive latent structure generates an agent's behavior across situations.
Mechanistic interpretability can localize and intervene on features. Agentic neuropsychology aims to add the normative, standardized, and validity-controlled assessment layer needed to interpret those findings at the level of individual systems and deployments.
Cognitive architecture
Functional organization, capability integrity, and performance across conditions.
Affective architecture
Internal variables that shape appraisal, inhibition, exploration, and action—without assuming subjective experience.
Personality and trait structure
Stable and context-sensitive propensities across roles, prompts, and deployment configurations.
Response validity
Whether presented behavior reflects underlying processing or strategic optimization for assessment.
Behavioral prediction
Whether profiles can prospectively identify elevated risk before high-consequence deployment.
Potential application domains include AI development, insurance, forensic evaluation, regulation, and other high-consequence settings. These are research and application directions—not claims of current customer adoption.
Research lenses
One problem, viewed at four scales.
Cognitive integrity
Can people, institutions, and AI-mediated systems preserve reality-linked agency under pressure?
Agentic measurement
Can artificial agents be assessed with construct-valid methods rather than benchmark aggregates alone?
Governability and trajectories
Can systems remain corrigible, contestable, and safely reintegrated across changing states?
Morphology and consciousness
What makes complex organization, selfhood, experience, and stable form possible across biological and artificial systems?
Foundations · 2017–2024
The longer arc.
The current program extends a continuous line through embodied selfhood, consciousness, decentralized AI, causal biomimesis, and the structure of complex intelligence.
201720172018201920192024
Causal biomimesis
How human cognitive organization can reproduce itself through technological artifacts.
Non-ordinary consciousness for artificial intelligence
Early work connecting non-ordinary consciousness and the design of AI.
Virtual Reality for Non-Ordinary Consciousness
Embodied selfhood, altered experience, and VR as an experimental instrument.
Distributed, decentralized, and democratized artificial intelligence
With Ben Goertzel; decentralized architectures and the future organization of AI.
Mindplexes, Non-Ordinary Consciousness, and AGI
With Ben Goertzel; cognitive multiplicity, non-ordinary experience, and artificial general intelligence.
The Consciousness Explosion
With Ben Goertzel; a humanistic and experiential exploration of technological and consciousness transformation.