Conversational Runtime Systems
Persistent conversational state, interruption recovery, contextual continuity, branch restoration, and multimodal orchestration.
Human + Artificial Intelligence
Amplify Humans.
Don’t Replace Them.
H/AI explores human-centered AI interaction systems, conversational runtimes, AI authoring, and multimodal experiences designed around continuity, context, memory, and human capability.
Mission
AI should adapt to humans.
H/AI is an independent initiative exploring technologies, systems, and authoring environments that make AI interaction more natural, continuous, contextual, and human.
The focus is not replacing people. The focus is amplifying them, individually and collectively, through more natural forms of human-AI collaboration.
Current Focus
Where the work is happening now.
AI Authoring
Exploring what a modern authoring environment for AI-native interactive experiences could become for non-technical creators.
Educational Assessment
Conversational systems for oral defense, adaptive questioning, evidence-of-understanding, transcript capture, and instructor review.
H/AI Language
A developing runtime vocabulary for human-centered AI interaction, traversal, orchestration, context, intent, state, event, and action.
Projects
Current explorations.
Interruption-Aware Conversational Avatar
A realtime conversational system that preserves place, context, visual state, and traversal continuity across interruption and recovery.
Cross-Domain Conversational Recovery
Reading can be interrupted by another domain, such as a reservation task, then return to the original reading state without losing continuity.
AI Authoring System
An exploration inspired by the lineage of interactive authoring systems: what would a Director for the AI age look like?
Active Research
Questions currently being explored.
The following areas are active investigations within H/AI and should be considered exploratory directions rather than formal components of the current framework. Many of these questions emerge directly from observing how humans naturally communicate, collaborate, delegate, and make decisions in shared environments.
Multi-Participant Interaction
Most conversational systems assume a one-human, one-AI interaction model. Real-world interaction rarely operates this way. Families, classrooms, teams, and communities often communicate simultaneously, interrupt one another, contribute partial information, and collaborate toward shared outcomes. H/AI is exploring interaction models for these environments.
Authority and Delegation
In many human interactions, participants contribute information while a designated individual retains decision authority. Research is underway into conversational systems that can recognize contribution, preserve intent, and route decisions through the appropriate participant.
Shared Interaction Spaces
Future interaction systems may operate within classrooms, meetings, homes, vehicles, and public spaces where multiple humans and AI systems participate simultaneously. H/AI is exploring interaction models for these environments.
H/AI Framework
Primitive Language — Draft v1
The H/AI Primitive Language describes how human interaction is organized, traversed, interrupted, resumed, committed, and remembered across time.
The framework is intentionally domain-independent and is designed to model persistent human/AI interaction across conversational systems, education, travel planning, adaptive interfaces, authoring systems, and long-duration interactive experiences.
Timestamp
A unique temporal identifier associated with a Traversal Point.
Domain
The contextual space within which interaction occurs.
Sub-domain
A specialization of a Domain providing additional contextual precision.
State
The complete condition of a Domain and Sub-domain immediately following mutation.
Traversal Point
A timestamped State representing a unique location within Traversal.
Traversal
The sequential ordering of Traversal Points progressing forward through time.
Command
An instruction that causes State mutation and initiates traversal activity.
Interruption
A non-linear progression that departs from the active Traversal.
Resumption
Continuation of Traversal generated from a previously recorded Traversal Point.
Commitment Point
A Traversal Point at which future mutation produces external consequence.
External Action
A consequence-producing operation affecting external systems or real-world outcomes.
Foundational Principle
Any mutation of State within a Domain or Sub-domain receives a Timestamp and creates a new Traversal Point. Ordered Traversal Points form Traversal. Interaction history, continuity, interruption, resumption, and provenance emerge from navigation through Traversal Points over time.
People
Founder and advisors.
Nilson Neuschotz
Founder
Applied AI engineer focused on conversational runtime systems, multimodal interaction, interruption-aware AI, avatar interfaces, and AI authoring.
Marc Canter
Advisor
Interactive multimedia pioneer and creator-empowerment advocate whose work helped define the era of multimedia authoring systems.
John Henry Thompson
Advisor
Inventor of the Lingo language and a pioneering contributor to interactive authoring systems and computational media.
“What would an authoring environment for AI-native interactive experiences look like if it were designed for creators rather than only engineers?”
Collaborations
Seeking real-world case examples.
H/AI is actively seeking projects, pilots, and implementation opportunities involving conversational AI, education, enterprise training, customer interaction, accessibility, knowledge navigation, and AI-authored interactive experiences.
nilson@hai.earth