🤖 About the Project
Validation methodology and AI transparency
Validation Methodology
Every claim from the podcast was validated following a rigorous and systematic process to ensure scientific integrity of this project.
Identify verifiable claims from the podcast
Identify underlying concept or mechanism
Search peer-reviewed studies on PubMed/PMC
Analyze methodology and effect size
Assign support level (✅/⚠️/❌)
Record nuances, limitations, and links
AI Transparency
This project uses AI as a research assistant, not as a source of truth. Here's exactly how we use it and maintain scientific rigor.
Literature Search
AI helps find relevant studies on PubMed/PMC, but every study is manually reviewed for quality and relevance.
Semi-automatedSummary Writing
AI assists in summarizing complex findings, but all summaries are fact-checked against the original research.
Human-supervisedNo Hallucinations
Every claim requires a link to an actual peer-reviewed source. We don't accept AI-generated citations without verification.
VerifiedValidation Rules
Strict rules prevent overconfidence: partial evidence is marked as such, and claims without evidence are clearly labeled.
ConservativeEvidence Levels
Not all evidence is created equal. We classify support using strict criteria.
Validated
Direct evidence from multiple peer-reviewed studies with strong methodology and consistent findings.
Partial Evidence
Some evidence exists but is limited, indirect, or inconsistent. More research needed for definitive conclusions.
Not Supported
No peer-reviewed evidence found, or evidence directly contradicts the claim. These claims are clearly labeled.
Core Principles
Scientific Rigor First
No alarmism, no exaggeration. If the science doesn't support it, we don't say it. Period.
Direct Sources
Every claim links to original peer-reviewed sources. Read them yourself and verify.
Nuance Matters
Science is rarely black and white. We document limitations, conflicting evidence, and contextual factors.
Open to Updates
Science evolves. When new research emerges, we update our validations accordingly.
Questions or Feedback?
Found an error? Have a suggestion? Want to contribute research? We value scientific accuracy above all and welcome corrections.
Contribute to the Project