Field Notes are practitioner-level observations from live IBM i AI deployments, architecture decisions, governance events, and activation signals. Not predictions. Not vendor analysis. Events that have landed and what they mean for organizations running the world's most resilient enterprise platform.
On May 19, IBM confirmed its participation in Project Glasswing — an industry coalition anchored by Anthropic. IBM Concert and IBM Autonomous Security are Power-native, hybrid cloud-native. IBM is building an agentic security layer around the same stack you operate on. The perimeter just moved.
ChatGPT's personal finance launch — 12,000+ institutions via Plaid — is not a consumer product story. It is the opening move in the agentic traffic era. The DB2 data is there. The RPG logic is there. The surface that lets an agent reach it isn't built yet. Four walls, one bridge, three questions to answer before I3.
31% of leaders are placing AI agents on their org charts as employees. BCG's randomized experiment shows why that fails. The naming error is a symptom — the bolt-on problem is the disease. Governance models built for employees produce agents that can't be governed.
Washington is building a real AI governance framework. It reviews models. It does not review deployments. For IBM i shops running agentic AI in production, the audit trail and SOX exposure are still yours to govern. The regulatory gap is a deployment gap.
The economics of all on-premises hardware have changed. The argument against IBM i was never about the platform — it was about hardware ownership. Hardware ownership just became indefensible for everyone. The IBM i read of a story unfolding at macro scale and landing directly on your infrastructure decisions.
IBM's Cyber Fraud Detection product analyzed through the Knowledge Distance mechanism. The 90% number reframed. Enterprise validation of the agentic investigation architecture at scale. IBM Think 2026 as signal confirmation — the product does exactly what the KD framework predicts an agentic layer must do.
The formal KD definition. Harvard and Stanford ran the experiment. Domain proximity is the binding variable determining AI output quality. Six root causes that all collapse to one. The IBM i structural advantage named precisely — data proximity, process proximity, platform proximity. The binding variable in every AI deployment failure.
The complete framework behind Field Note 04. Four independent threads — Naval Ravikant on specific knowledge, Wittgenstein on private language, operator-level observation, and economic analysis — converging on the same finding. The Harvard/Stanford GenAI Wall mechanism, the three-state transformation model, and the IBM i structural positioning argument at full depth. Published on reggiebritt.ai.
The Wall Street Journal named the most valuable new role in the AI economy. Aaron Levie called it the internal FDE. The IBM i practitioner has been performing it for thirty years without the title. WSJ data, Cloudflare, Block, Amazon, and COMMON's educational path — the full signal mapped to the practitioners who already hold this role.
The IBM i practitioner is navigating the loudest transition in enterprise technology without a map. Here is the argument that they have been holding one the whole time. Knowledge Distance. Sovereignty. The Human-Agentic destination. Made as a declaration, not a roadmap.