vibeweaving
Research Lab · current testbed: rules-based textile generation

Designing friction that changes judgment.

Vibeweaving Lab tests how interface rules affect creative judgment. The generator stays the same while the interaction changes. We track what people notice, revise, accept, delay, or refuse.

Lineage Canvas: Productive Friction in Born-Digital Heritage. Expanded 2026, ACM.

Open Governed Studio Read the art paper
Rule-based browser instrument No cultural-dataset training No runtime model inference Weaving is the first testbed, not the limit
Interaction Wind Tunnel / condition preview
Capability held constant
Source
Documented research reference
Action
Transform
Engine
Rule-based construction
Request
Generate a new visual field
→
Interaction condition
→
Available exit
What the study observes

01 / Research question

When does friction support reflection?

Friction is not useful by default. We compare specific interventions and measure what they change.

The generator stays the same. The interaction changes. We track reliance, revision, agency, and judgment.

What stays fixed
The underlying generation capability, task, source scenario, and output space.
What changes
Timing of questions, inherited restrictions, visible consequences, review holds, and refusal.
What we measure
Revision paths, reliance patterns, stopping points, confidence, perceived agency, rationale quality, and expert assessment.
Research boundary
The browser demonstration uses generic source classes. Access, documentation, or field conversation never produces commercial permission.
02 / Research instruments

Three scales of governed generation.

The same research question is tested in a browser tool, a spatial interface, and a public installation.

03 / Core method

Interaction Wind Tunnel

The Wind Tunnel compares four interaction conditions while the task, source, and generator stay the same.

Same task · same source · same generation engine Only the interaction varies

Intervention
    Available exit
      Measures
        These are study conditions, not results. Participant data is required before we can claim an effect.
        04 / Research evidence and outputs

        Working instruments and study materials.

        This section separates functioning instruments, study protocols, records, and proposed public work.

        05 / Work with the Lab

        Test the method in a real institution.

        We work with museums, educators, researchers, and creative practitioners on pilots, workshops, and comparative studies.