Most AI deployments today are optimized for the wrong problem. They compress existing documentation, accelerate existing workflows, and surface existing knowledge faster. The harder problem — the one almost no one is solving — is capturing reasoning that was never documented in the first place, from carriers who are still alive but won't be for long, in a form that survives the carrier and remains queryable by people who were never in the room.

This session presents a working architecture for that problem.

The historical case is structural: for four hundred years, the most catastrophic loss in Black America was not economic. It was the systematic destruction of the intergenerational knowledge transfer infrastructure — the mechanism by which one generation's hard-won reasoning became the foundation the next generation built on. Event knowledge survived. Reasoning knowledge did not. Sixteen generations of restarted intelligence, because the transfer mechanism — human memory, human voice, human presence — had a fatal constraint: remove the carrier and the deposit goes with them.

That constraint is now solvable. Not metaphorically. Technically.

The talk walks the audience through three layers of working architecture: (1) the excavation protocol — a structured interview methodology that surfaces governing doctrine, not just stories, from a living carrier; (2) the deposit format — how raw excavation output is encoded into a governed, queryable reasoning artifact that survives the carrier; and (3) the deployment layer — how the resulting system is made sovereign, permissioned, and accessible to authorized recipients across time, without leaking into the broader model substrate.

This is not a thought-leadership talk dressed up in technical language. It is the system the speaker is actively building and shipping, presented with the engineering decisions visible — what was tried, what failed, what the current architecture commits to, and where the open problems still are.

Audiences leave with a working mental model of reasoning-knowledge capture as a deployable AI category, a clear-eyed view of where current LLM and RAG approaches fall short of it, and a decision: name the carrier in their own life or institution whose reasoning has not yet been deposited, and the deadline by which it will be.

First public delivery. Platform debut of the architecture developed in Ant Blair's third book, The Unfinished Inheritance: How AI Restores the One Thing Slavery Destroyed That No Policy Ever Addressed (Blade & Signal Publishing, 2026). Booking partners receive a delivery that has been pressure-tested in writing across 240 pages and is being shipped as live production infrastructure in parallel.

Target audience. AI practitioners, engineering leaders, applied-AI product teams, knowledge management architects, CTOs and CIOs at institutions with high-stakes institutional memory loss risk, and conference programs covering AI applications, AI ethics, governed AI, knowledge graphs, retrieval architectures, and AI for cultural and civic infrastructure. The talk also rewards mixed audiences where technical leadership sits alongside executive sponsors — the architectural argument is technically substantive and strategically actionable in the same delivery.

Topic categories. Applied AI / Generative AI / Governed AI / Knowledge Architecture / Human-AI Interaction / AI Ethics / AI for Civic and Cultural Infrastructure.

Preferred session duration. 45 minutes for the keynote. 60 minutes with Q&A at the platform. Workshop variant available at 90 minutes (see below).

Format. Single speaker, platform keynote. No panel format — the architectural argument requires a sustained line of reasoning that panel structures break.

Workshop variant. A 90-minute hands-on session can be added in which attendees walk through a live excavation protocol on a sample carrier and observe the deposit-encoding step in real time. Requires room configuration for facilitated work (no fixed AV requirement beyond standard breakout).

Expandable engagement — The Unfinished Inheritance Experience. The keynote is the entry point to a three-part engagement available to host organizations: (1) platform keynote, (2) facilitated town hall on the institutional implications, and (3) a knowledge excavation workshop that produces a deployable reasoning deposit from a named carrier in the host community before the speaker leaves the venue. Useful for conferences with a co-located executive track or institutional partner program.

Technical requirements. Wireless lavalier or headset microphone. Confidence monitor preferred. Single screen, HDMI input, 16:9. Standard keynote configuration — no special staging. Speaker travels from Flint, Michigan, USA.

Credentials. Seven-time public speaking champion. TEDx speaker (The Access Code, TEDx Bloomington, 108,000+ views). Founder of the first African-American-owned digital marketing agency in Michigan. Author of three published books. Currently building production AI infrastructure for governed reasoning capture and intergenerational deposit.