AI Visibility Field Notes
Short, real world observations documenting how AI systems interpret identity, authority, and topical focus in practice, based on live operator decisions and their downstream effects on AI visibility over time.
Additional, field notes for practical fixes for AI Visibility can be found here: https://josephmas.com/ai-visibility-field-notes/ai-visibility-field-notes-practical-fixes-for-observed-issues/
-
Continuity thinning across ChatGPT update cycles
Posted by: Joseph MasDocument Type: AI Visibility Field Note Observation Across multiple ChatGPT update cycles, long run continuity appears thinner than it was previously. The most noticeable change is not single turn accuracy. It is the way longer narrative threads reset sooner and require more restating of prior context. What appears to be changing Long Read more
-
AI Visibility Field Notes: Practical Fixes for Observed Issues
This page contains multiple independent entries. Entries are separated by structured headings and ordered chronologically. Document type: AI Visibility Field Note AI Visibility Field Notes documents structured operational records of AI system behavior patterns and corrective implementations. Each entry documents an observed issue, the corrective action applied, and the resulting change. Field Notes Field Note: Read more
-
Markdown Pages for Google versus LLMs
By Joseph MasPublished: 11/25/2025Revised: 1/6/2025Document type: AI Visibility Field Note Observation Recent discussion around LLM only Markdown pages highlighted differences between how Google Search and large language models ingest and interpret content for recall, following comments by John Mueller in a thread raised by Lily Ray. Context The discussion emerged from commentary on whether Markdown Read more
-
Clean Data Beats Clever Prompts
By Joseph Mas Published: 11/23/2025Revised: 1/6/2026Document type: AI Visibility Field Note AI Visibility Field Notes This field note documents why clean structured data functions as the primary constraint for reliable AI ingestion, and why downstream prompt techniques cannot compensate for weak or ambiguous data foundations. Observation Across multiple generations of search and retrieval systems, cleanly Read more