Find the moments that shape an interaction.
Temporal Scope reveals recurring patterns in human–AI interaction. Select any moment, find structurally similar occurrences, and examine how coordination, drift, breakdown, and repair unfold across time.
A product-first instrument for observing, searching, and comparing temporal patterns.
We can measure AI outputs. We still struggle to understand the interactions that produce them.
Organizations evaluate AI through accuracy, task completion, latency, ratings, and aggregate metrics. But when a collaboration succeeds or fails, those measures rarely reveal where the interaction changed, whether the same pattern has occurred before, or why the system recovered in one case but not another.
Existing tools tell you:
- whether the task succeeded;
- how the user rated it;
- how long it took;
- how many corrections occurred.
Temporal Scope helps reveal:
- where coordination began to deteriorate;
- whether the pattern is isolated or recurring;
- what typically precedes it;
- what distinguishes repair from continued breakdown;
- how human and AI trajectories shape one another.
Temporal Scope solves the problem of temporal invisibility in human–AI systems.
From observation to explanation.
See
Identify a moment of coordination, hesitation, drift, breakdown, repair, or emergence.
Select
Drag across the interval directly on the temporal chart.
Search
Find structurally similar occurrences across sessions and datasets.
Explain
Compare their context, interaction dynamics, outcomes, and probable continuations.
Discovery begins before the phenomenon has a formal label.
See something meaningful? Search for its relatives.
The analyst does not need to define the phenomenon in advance. The selected interval becomes the query.
Explore the Temporal Scope prototype.
The live browser-based prototype includes two instrumented drawing interfaces, the Creative Trajectory Laboratory and Aether: Enactive Co-Creative Drawing AI. Researchers can use the built-in drawing prototypes or upload their own CSV for analysis, which includes Human and AI signal modes, Local–Regional–Global attractors, regime formation, drift pressure, future-state predictions, video-grounded coding, temporal structure reports, motif libraries, and query-by-example search.
The prototype runs locally in the browser; its current interface describes itself as an open research platform and notes that uploaded data does not automatically leave the browser.
Available in the live prototype
Temporal discovery: motifs, regimes, transitions, attractors, recurrence, drift, and adaptive forecasts.
Interaction analysis: aligned Human and AI trajectories, coupling, participation, and predicted next states.
Evidence workflows: synchronized video coding, expert verification, correction, annotation, and corpus export.
Reports and libraries: temporal structure reports, motif export, and reusable motif collections.
Similar shapes do not always mean similar interactions.
A generic motif-search system says, “These curves look alike.” Temporal Scope examines how the participants relate, what preceded the pattern, where their trajectories diverged, and what happened afterward.
Occurrence A
The human becomes uncertain. The AI detects the shift and changes strategy. The interaction recovers.
Occurrence B
The human follows a similar trajectory. The AI continues its prior strategy. Misalignment increases.
Interaction microscope plus search engine.
The microscope
Inspect raw trajectories, regimes, recurrence, coupling, drift, and predictions across multiple timescales.
The search engine
Use any selected temporal interval as a query and retrieve structurally related moments.
The explanatory layer
Compare local, regional, and global structure, preceding context, subsequent outcomes, and alternative continuations.
The chart lets you see the phenomenon. Search finds its relatives. Temporal analysis helps explain the difference.
Explore these capabilities in the live prototype ↗
Initial high-value use cases.
Diagnose interaction breakdowns
Find recurring patterns of misalignment and identify the point where the interaction begins to deteriorate.
Discover successful repair
Compare effective and ineffective recovery attempts after misunderstanding or drift.
Evaluate models and interfaces
Compare not only final outcomes, but the trajectories through which different systems reach them.
Understand creative collaboration
Identify patterns of convergence, initiative transfer, fixation, divergence, and co-creative flow.
A concrete example.
A user begins repeatedly reformulating the same request while the AI continues producing increasingly detailed versions of the same mistaken interpretation.
Temporal Scope finds 11 related interaction episodes.
Successful recovery was not associated with a longer answer. It occurred when the AI changed the interactional strategy rather than elaborating its existing interpretation.
Built for teams studying how intelligence works together.
AI product teams
Understand repeated friction, correction, disengagement, and successful adaptation.
Human–AI researchers
Discover phenomena, construct operational definitions, and compare interaction trajectories.
AI evaluation and assurance teams
Investigate whether problematic episodes are isolated, recurring, escalating, or repairable.
Collaborative and creative AI teams
Study co-creation, initiative, alignment, convergence, and interaction quality.
What a Temporal Scope Pilot Includes.
A pilot is a focused engagement around one real human–AI workflow, one recurring interaction problem, and a bounded body of temporal evidence. The goal is to produce findings that can directly inform research, product design, or evaluation.
1You provide
- repeated human–AI interaction records;
- one concrete question or recurring problem;
- a domain expert who can help interpret the findings.
2We provide
- data preparation and Temporal Scope analysis;
- recurring-pattern and meaningful-variant discovery;
- breakdown-versus-repair comparison;
- an annotated interaction-pattern library;
- a findings presentation with design recommendations.
3Typical pilot
- 4–6 weeks;
- one bounded workflow or research question;
- private or local analysis where appropriate;
- a final analytical report and review session.
Built for exploratory human–AI system evaluation.
Temporal Scope is currently being developed as a research and evaluation platform. We are seeking pilot partners with repeated human–AI interaction data and a concrete need to investigate breakdown, repair, adaptation, or collaboration quality.
Start with a bounded question
A strong pilot begins with one repeated workflow and one question: where does interaction quality change, which patterns recur, and what distinguishes better outcomes from worse ones?
Nicholas Davis, PhD
Human-centered computing researcher developing Enactive AI, Interaction Science, and computational instruments for studying cognition, creativity, and collaboration through time.
A theory of cognition, creativity, and co-creation developed since 2014.
Temporal Scope grows from a long-running research program centered on a simple claim: cognition is not adequately understood as isolated information processing inside an individual system. Cognition develops through embodied activity, temporal continuity, environmental structure, and ongoing participation with other people, tools, and intelligent systems.
Within this view, creativity is not merely the production of novel outputs. It is an evolving process in which possibilities are generated, selected, stabilized, revised, and transformed through interaction. Human–AI co-creation therefore depends on more than accurate prediction. It depends on whether the participants can remain responsive to one another, recognize drift, repair misalignment, preserve productive difference, and reorganize the collaboration when its current pattern stops working.
Since 2014, this program has developed through work on enactive cognition, creative sense-making, Interaction-Centered Intelligence, co-creative AI, enactive AI, Enactive Drift Regulation, and the design of temporal instruments capable of making these dynamics observable.
An instrument for a temporal science of interaction.
Temporal Scope begins with human–AI interaction, but its larger purpose is to study how meaningful patterns form, persist, recur, drift, coordinate, and reorganize through time.
Temporal Science
Study of organization and transformation through time.
Interaction Science
Study of relational organization across trajectories and timescales.
Temporal Scope
An instrument for observing, searching, and comparing temporal patterns.
Find recurring patterns of success and failure. Compare their contexts. Understand what changed.
Design systems that respond more effectively through time.