Edition 01
Reflection over completion
The Cognitive Mirror
An AI that makes the shape of thought visible—separating facts, assumptions, feelings, and values before offering an answer.
What am I really saying?
Designing the Relationship
Between Humans and AI
Notes on agency, understanding, memory, and everyday intelligence.
A working thesis
AI can do more than complete a thought. It can help us see how we think, build richer models of the world, and carry understanding forward over time.
Issue 01 · A thesis in three forms
Beyond the answer
Three roles for AI that shift the interface from output to understanding.
Edition 02
Models over answers
The Co-Modeler
An AI that builds a living explanation with us: causes, incentives, feedback loops, tensions, and missing evidence.
What are we building together?
Edition 03
Continuity over chronology
The Thinking Space
An external space where conversations become a connected knowledge graph—growing across days, topics, and contexts.
What connects over time?
Recent inquiries
Notes from the field
Each piece is one observation in a larger, continuously evolving line of inquiry.
Interaction NoteObservation
Chapter 01
When reflection becomes another kind of interruption
A close reading of the moment an assistant names a feeling before the user is ready to name it.
07.18.266 min read
Product TeardownTeardown
Chapter 03
Voice is not more human yet
Latency disappeared. Turn-taking improved. Why do many voice experiences still feel less considerate?
07.04.269 min read
Human–AI EssayReflection
Chapter 04
Who owns the model of me?
Memory is not only a retrieval problem. It is a question of authorship, legibility, and the right to revise.
06.22.2612 min read
Monthly SignalsResearch Note
Field Notes · 06
Six signals from everyday AI
From answer fatigue to prompt rituals: small behaviors that hint at a larger change in how people delegate thought.
06.01.268 min read
Concept ProposalProduct Proposal
Chapter 02
From answers to inspectable thought
A product mechanism for turning one response into a model that users can challenge, extend, and make their own.
05.19.2610 min read
Human–AI EssayResearch Note
Chapter 02
The conversation is the wrong container
Timelines preserve what was said. Thinking spaces should preserve what changed.
05.02.267 min read
The collection
One growing narrative
The archive is organized as a four-part inquiry—moving from understanding, to thought, to action, to relationship.
How AI understands people
Listening · Context · EmotionWhat changes when listening means holding ambiguity—not rushing to infer intent?
How AI organizes thought
Reflection · Models · MemoryCan an interface help us see how we think, not simply accelerate what we produce?
How AI enters everyday action
Voice · Agents · AutonomyWhere should assistance end, agency begin, and responsibility remain visible?
How AI changes our relationships
Self · Others · SocietyWhat happens to the relationship with ourselves—and with others—when intelligence is always present?
Five ways into the workA research archive,
A research archive,
not a content feed.
Every entry carries both a chapter and a format. Together they reveal not just a list of articles, but a research practice taking shape over time.
Interaction Notes
Close readings of real human–AI conversations and the micro-interactions hidden inside them.
Product Teardowns
How voice, memory, search, agents, and emotional responses actually feel in use.
Human–AI Essays
Longer inquiries into agency, understanding, autonomy, and the relationships we are forming.
Field Notes / Monthly Signals
A running record of product shifts, emerging behaviors, and signals worth watching.
Concept Proposals
Research translated into product principles, mechanisms, and testable MVPs—not speculative screens.
About this journal
The most important question is not how intelligent AI becomes, but what kind of relationship its intelligence makes possible.
Relational Intelligence is an independent design research journal about the space between people and intelligent systems.
It studies everyday interactions, traces product signals, and turns research into principles and mechanisms that can be tested.