Open AI Search Research Lab

Badges

Status and evidence are set per experiment in frontmatter (status, evidenceLevel). Raise evidence only when archives justify it: D → C → B → A. Never invent upgrades.

Canonical sources

Preferred URLs from barbu.es (also listed in llms.txt ).

• About https://barbu.es/about/
• Cornerstone https://barbu.es/writing/founder-how-i-think/
• Now https://barbu.es/now/
• Work https://barbu.es/work/
• Experience https://barbu.es/experience/
• Writing https://barbu.es/writing/
• Contact https://barbu.es/contact/
• Resume https://george.barbu.es/
• Portfolio https://portfolio.barbu.es/
• InfoWebPlus https://infowebplus.com/
• llms.txt https://barbu.es/llms.txt
• LinkedIn https://www.linkedin.com/in/barbugeorge/
• GitHub https://github.com/george-barbu-es

Standards & validators

Primary sources (for example schema.org ) keep claims auditable.

1. Exp-001 Does llms.txt influence AI discoverability? Running C

   Does publishing a well-formed /llms.txt file change how major AI systems discover, cite, or summarize a controlled web property compared to an otherwise identical baseline without llms.txt?

2. Exp-002 Does adding Person + Organization schema improve AI understanding? Running C

   Does deploying coherent schema.org Person and Organization structured data on a canonical site improve AI systems' accuracy when describing the person-organization relationship, compared to a markup-absent baseline?

3. Exp-003 Can supporting entities strengthen a canonical organization? Planned D

   Does deliberately publishing and cross-linking a supporting entity chain (George Barbu → barbu.es → InfoWebPlus → client websites) increase AI systems' correct recognition of InfoWebPlus as the canonical organization hub?

4. Exp-004 Entity stacking using legitimate Technology Partner pages Planned D

   Do transparent Technology Partner / attribution pages on third-party sites improve discoverability and correct entity association for the credited vendor, compared to sites without such pages?

5. Exp-005 Can AI citations be influenced through high-quality technical documentation? Planned D

   For the same underlying product or organization, do AI systems cite high-quality technical/research documentation more often and more accurately than marketing content?

6. Exp-006 How do different LLMs interpret the same organization? Planned D

   Given identical probe prompts about one organization, how do ChatGPT, Claude, Gemini, and Perplexity differ in factual coverage, mutual agreement, hallucinations, and source behavior?

7. Exp-007 Can free tools create stronger entity associations than blog articles? Planned D

   Does publishing a useful free tool on a controlled domain create stronger AI entity associations (brand ↔ capability) than publishing a conventional blog article on the same topic?

8. Exp-008 Does author consistency improve entity confidence? Planned D

   Does maintaining consistent author identity signals (same photo, bio, titles, and company) across multiple domains increase AI systems' confidence and correctness when answering questions about the author-organization relationship?

9. Exp-009 Testing structured data combinations Planned D

   Which combinations of schema.org types (Person, Organization, CreativeWork, SoftwareApplication, Article, ProfilePage, CollectionPage) produce measurable differences in AI entity interpretation and citation behavior on a controlled site?

10. Exp-010 Can real case studies outperform generic SEO articles for AI citations? Planned D

    Do detailed, real case studies receive more accurate AI citations and stronger brand attribution than generic SEO articles targeting similar queries?