Build In Public
How Cannabinoid Research Transparency Should Work
Use this page as the public operating standard for the site. The encyclopedia should teach first, show sources, invite correction, publish quality context, and keep education separate from sales pressure.
What to know first
Three ideas that make this easier to judge
The roadmap for turning Cannabinoid Encyclopedia into a public education system with traceable sources, correction pathways, and transparent uncertainty.
Key idea
Education is the first CTA
Readers should be able to learn, search, verify, and follow without being pushed toward a conclusion.
Key idea
Corrections are a feature
A living evidence site earns trust by making source suggestions, expert review, and updates visible.
Key idea
Quality claims need context
Sourcing, testing, formulation, and COA language should be explained as quality context, not health-outcome proof.
Apply the standard
How to inspect the supporting record
Application check
Make every research statement traceable
A reader should be able to move from plain-language prose to the exact evidence row, source record, and relevant passage or structured locator. The source trail should preserve compound, formulation, dose, route, population, study design, outcome, and limitations rather than presenting a citation as decoration.
Application check
Publish corrections as part of the record
Updates should identify what changed, why it changed, which source or review prompted the revision, and when the new version became public. Prior assessments should be superseded or versioned instead of silently overwritten so that editorial decisions remain inspectable over time.
Application check
Separate education, review, and commercial influence
Evidence pages should disclose relevant conflicts, keep AI output draft-only, and require human approval for public claims. Sponsorship, products, or commercial relationships must not determine evidence classification or turn an educational conclusion into individualized advice or a product recommendation.
Application check
Publish null and insufficient findings
A transparent evidence system cannot select only positive results. Null findings, conflicting outcomes, absent data, and conclusions of insufficient evidence belong in the public record when they answer the review question. The same inclusion rules should apply regardless of whether a result supports an appealing headline or commercial narrative.
Application check
Show where automation stops
Software can search, organize, validate, and draft, but it should not approve its own research claim or publish autonomously. Public methodology should identify the human review gate, the evidence state required for publication, and the checks that keep mechanistic, animal, observational, and controlled human findings in separate lanes.
Application check
Define ownership and conflicts
Every correction, review decision, and approved page should have an accountable human or editorial role. Relevant financial, research, advocacy, or product relationships should be disclosed where they could affect interpretation. Governance is part of the evidence trail because readers need to know both what supports a statement and who approved it. Access roles, review dates, and the reason for a decision should be recorded consistently enough for a later editor to reproduce and audit the publication state.
A simple checklist
What to look for
- 1
Grow the research graph. Keep expanding compound, target, outcome, safety, dictionary, and guide coverage.
- 2
Publish quality literacy. Add COA explainers, sourcing documentation routes, formulation standards, and transparent caveats.
- 3
Keep education distinct. Any non-educational layer should stay clearly separated from the encyclopedia's source-backed education pages.
Important limits
What this information cannot prove
- 1
Do not make the encyclopedia read like promotional copy.
- 2
Do not let any CTA outrank safety or evidence context.
- 3
Do not present educational pages as individualized medical advice.