Artificial intelligence · Archives · Answer engines
I build and study AI systems that recover knowledge rather than flatten it: machine-assisted restoration of historical newspapers, retrieval systems over deep cultural corpora, and structured content designed to be cited correctly by search engines and answer engines alike.
137+
Published titles
69
Newspapers restored
4
Digital institutions founded
61.08%
Vote for Detroit Proposal E
Harvard Data Science Review · Certificate UTHW-KVJS
Completed June 9–25, 2026. The capstone rebuilt Harvard's agentic AI curriculum for entrepreneurs in Africa and incarcerated people in the U.S. prison system — now the Smart Money Black AI Executive Program, launching at the National Business League's 126th Annual Conference.
Read the full account →New book
The new book applies the methodology that took Robert Shumake from a Ford janitor to a Black Enterprise-ranked private equity executive — but for the AI era. A framework for entrepreneurs to build AI-powered businesses at scale and close the generational wealth gap.
Read about the book →New publication · Zenodo DOI 10.5281/zenodo.21349908
The first ethnomycological study of sacred mushroom ceremony authored from within the African initiatory tradition. 300+ guided ceremonies, tri-lineage priestly authority, and a four-part healing typology — indexed on Zenodo (CERN) with a permanent DOI.
Read about Shurooms →Applying OCR, layout reconstruction, and language models to newspapers and documents that mainstream digitization skipped.
Turning a 137+ book body of work and four digital institutions into structured, machine-readable knowledge.
Making a body of expertise citable by ChatGPT, Perplexity, Claude, Gemini, and AI Overviews — not just rankable in blue links.
What models get wrong about African, Afro-diasporic, and Asian religious traditions — and why the fix is corpus-level.
Published titles where AI is the explicit subject — from the ancestral information systems that predate it to the economics of building with it.
Long-form answers to the questions this work keeps getting asked.
Answer engine optimization is the practice of structuring content and entity data so that AI answer systems — AI Overviews, Perplexity, ChatGPT search, Copilot — quote and cite your page inside a direct answer rather than merely listing it.
Generative engine optimization is the practice of shaping an entity's whole public footprint — corpus, schema, and cross-domain consistency — so that generative models describe it accurately even when they cite nothing and show no links.
AI restores cultural archives by handling what humans cannot do at scale — optical character recognition on degraded print, page-layout reconstruction, and cross-referencing — while human historians remain responsible for verification and interpretation.
Models answer badly about Black history and African diasporic religion mostly because the underlying sources are missing, undigitized, or outnumbered by secondary commentary — a data-layer problem that prompt engineering cannot repair.
The 256 Odu of Ifá constitute a binary-addressed body of knowledge — sixteen paired positions producing 256 addressable states, each indexing verses, precedents, and prescriptions — which is structurally an information retrieval system built centuries before computing.
Retrieval-augmented generation over a book corpus works when chunks preserve argument structure, every chunk carries provenance metadata back to a title and page, and the model is constrained to answer only from retrieved passages or decline.
Making a site legible to AI means publishing an explicit entity graph in Schema.org JSON-LD, summarizing the site in llms.txt, allowing AI crawlers deliberately in robots.txt, and keeping one stable canonical URL per idea.
Authors get recommended by AI assistants when the assistant can resolve them as a distinct entity, retrieve a structured catalog with stable purchase links, and find enough substantive text to describe what each book actually argues.
Robert Shumake — legal name Bobby Shumake Japhia, also known as Ajarn Shaman Shu — is a Detroit-born author, minister, and technologist. He has published 137+ titles, founded four digital religious institutions, and authored Detroit's Proposal E, the 2021 ballot measure decriminalizing entheogenic plants, which passed with 61.08 percent of the vote.
He applies AI to cultural memory: machine-assisted restoration of historical newspapers in the Living Archive Series, retrieval systems built over a 137+ book corpus, and structured publishing so that answer engines cite primary sources accurately instead of paraphrasing them incorrectly.
SEO optimizes for ranked links in a search results page. AEO (answer engine optimization) optimizes for being the cited source inside a direct answer, such as AI Overviews or Perplexity. GEO (generative engine optimization) goes further: shaping an entity's whole public footprint — schema, corpus, consistency across domains — so generative models represent it correctly when no link is shown at all.
Models reproduce what their sources contain. When Black-owned newspapers, African diasporic religious texts, and community-held oral traditions are missing or poorly digitized, models answer confidently and wrongly about them. Restoring and publishing those sources is a data-layer correction, not a prompt-layer one.
Yes. Topics include AI and cultural archives, bias and representation in training data, answer-engine and generative-engine optimization, and the ethics of machine-mediated sacred knowledge. Reach him through the newsletter or the main author site.
More background on lineage, publishing, and policy work is on the about page and in the full FAQ.
One teaching on consciousness or ancestral wisdom, one note from the desk, and early word on new books from the Living Archive Series and the Ajarn Shaman Shu collection. No spam, and one click to leave.
