78,326 observations · 219 countries · 30 indicators
Every prior technology widened access to consuming knowledge, to read, learn, and reach what others discovered. None widened access to producing it. Can AI become the first scalable channel that helps people cross that boundary?
Author. AI-assisted, human-directed.
Maps knowledge access across a six-level depth ladder (L0 basic literacy through L5 producing new knowledge) using a corpus of 78,326 observations across 219 countries and 30 indicators, benchmarked against UN SDG 4.
World-average access falls from 82.5% at basic literacy to 0.14% at the research frontier, roughly a 270x decline. The rich-poor gap is under 2x at literacy and roughly 75x at the frontier. About 99.86% of humanity only consumes knowledge; roughly 0.14% ever reaches where it is produced.
Framed the question, defined the depth ladder and indicator set, directed the synthesis, and wrote the paper.
AI-assisted data synthesis and drafting under direct review.
Working paper, version 1.1. Not yet peer reviewed.
Indicator coverage varies by country, and the depth ladder is a constructed framework rather than a standardized international metric.
This is the clearest statement of why I do this work: Bucket Foundation, early research-methods writing, education reform, and the current interest in AI-assisted scientific discovery are all downstream of this question.