What it owns
- Preference axes and stimulus generation
- A/B or multi-choice experiment policy
- Outside option and re-roll behavior
- Client design DNA and confidence
- Preference update over time
P06 · Experience knowledge node
Design taste and preference learner
Learn a client’s visual preferences through a minimal sequence of high-information comparisons, separate from choosing individual components.
Problem boundary
What this subsystem owns and what it must hand to the rest of Actionist.
What research already found
- Prior research proposed contrast, radius, shadow, border, typography, density and chroma as mechanical knobs
- Sequential galleries and Bradley–Terry/Luce models are relevant
- Forced choices without “none” pollute the preference model
Open first-principles questions
- Is the mathematical minimum 7, 14 or another number of choices?
- Which axes are independent?
- Should choices use whole screens or controlled fragments?
- How does preference confidence change by component context?
Inputs
- Design stimuli
- Client choices
- Brand constraints
Outputs
- TasteProfile
- PreferenceConfidence
- DesignDNA
Connected knowledge nodes
Scientific research protocol
Every research owner uses the same evidence funnel. No padding is allowed when a denominator is inapplicable or evidence is unavailable.
Prior evidence auditRead the canonical spine, all domain-linked artifacts, the 17-industry catalogue and existing contradictions before searching again.
Commercial 100 → 10Define a denominator, map approximately 100 companies/products, and produce first-party dossiers for the strongest 10.
GitHub 100 → 10Search and dedupe approximately 100 relevant OSS projects, then deeply inspect the strongest 10 at the capability and architecture level.
Local estate joinInspect AutoSaaS, Great Library, SISOCRM, 21st stores and other domain-specific precedents already on the laptop.
Innovation 100 → 10Generate a wide divergent idea register, score it against the problem and retain the 10 strongest falsifiable innovations.
First-principles synthesisDecompose the irreducible problem, challenge inherited assumptions, reconcile contradictions and state the minimum sufficient design.
Contracts and experimentsReturn the proposed contract, dependency edges, unknowns, falsifiers, experiments, stop rules and decisions.
Independent verificationA separate verifier checks counts, links, source claims, overlap, contradictions and unsupported conclusions before promotion.
Evidence and agent outputs
Existing research remains source evidence. New agent packets are written to one owned workstream directory and linked here after verification.
| Artifact | Location | State |
|---|---|---|
| UI pick-to-spec research | research/ui-pick-to-spec-2026-08-27.md | existing |
| Token-pack science | research/token-pack-science-2026-08-27.md | existing |
| Taste picker demo | Open public artifact ↗ | existing |
| Future verified agent packet | research/workstreams/P06-design-taste-and-preference-learner/runs/<run-id>/ | planned |