The gap between what your organization knows and what your systems can act on. It is the binding constraint on agentic readiness — the reason AI stalls even when the technology works — and it is the one number the diagnostic is built to read.
Your organization already knows how it runs — decades of it, encoded in code and held in practitioners' judgment. Knowledge Distance is how far that knowing sits from what an agent can actually act on. Close it, and proven logic becomes an agent's foundation. Leave it open, and more AI activity just ships degraded output, faster. It is, precisely, the width of the seam.
Every other readiness problem is downstream of this one. You can buy the models, run the pilots, and stand up the infrastructure — and still fail, because the agent is acting across a distance it was never given the means to cross. The capability is rarely the limit. The distance is.
IBM i organizations start with much of the distance already closed. The logic is written down — journaled, secured, correct. The hard part is already built. But the distance is not yet zero, and the work of agentic readiness is, precisely, the work of closing it.
The diagnostic reads the distance across the HOT spine — Human, Organization, Technology. Each is a place the distance opens, and each has an approach to closing it. The Organizational axis carries the most hidden anatomy: it opens along four sub-dimensions, and most organizations are aware of only one.
The human dimension is the one everyone measures and few fully close — because it has two halves, and they are not the same work.
The technical capability the agentic era demands — agentic tooling, MCP fluency, the documented skills that make a practitioner current. Real, and increasingly well served by the IBM i community's own learning paths.
Verification, discernment, the human-change work that lets people — not just systems — achieve lift. The tacit knowledge that was never written down. This is the layer most readiness models underweight — and the one we address directly, through the NeuroChange Solutions method, delivered by certified practitioners.
That second half is the part that's easy to overlook — and the most overlooked source of readiness risk. A platform can be legible and an org can be restructured, and the launch still falters if the people can't hold judgment over what the agents do. It's the layer we watch most closely.
An organization can know more than it can articulate, share, or act on. The diagnosis is ours; the transformation runs on a proven model — ExO 3.0, delivered through certified practitioners.
The gap between what the organization knows and what it can express to an AI system. Tacit knowledge that lives in experienced practitioners — never documented — doesn't exist for the model. If it can't be prompted, it can't be used.
Egos, territorialism, and incentive misalignment actively prevent knowledge from surfacing — even when leadership wants it to. The org knows more than it will share.
Broken handoffs, undocumented decisions, and siloed operations that were survivable at human speed become critical failures at agentic speed. The org chart becomes transparent under load.
The gap between what the underlying technology can do and what the vendor has packaged it to do. The product is designed to sell; the capability behind it is often larger, but invisible through the commercial layer. This is the one most organizations never see.
The platform is deep, but it has no presence an agent can navigate by default — no declared policy, no machine-readable description, no governed path to act. This is the legibility distance: the layer between encoded business logic and the agent that needs to consume it. The work is a stack, built in order — discovery, comprehension, access, action, governance.
The Agent-Readiness Stack →Harvard Business School and the Stanford Digital Economy Lab ran a randomized experiment at IG, a UK trading firm — 78 employees, three groups at increasing distance from a task, with and without a bespoke GenAI tool. They were measuring something adjacent to what we measure: how far a person sits from the work, and whether AI can close it.
The finding is the one that matters here. AI closed the gap on conceptualizing the work — and hit a wall on doing it.
That second column is the whole argument. The AI produced usable work; the distant outsiders discarded the parts they didn't recognize as valuable. One participant, describing what he did with the tool's output:
I didn't fully understand what it was doing, because I never wrote an article like that. I added random stuff to make it more "marketing," you know what I mean? Data scientist · post-experiment interview
The model was never the constraint. The distance was. An agent's output is only as good as the judgment that can evaluate it — and that judgment is exactly what sits on the far side of the distance. This is why closing it is the work, and why the practitioner who holds the encoded logic isn't who agentic AI replaces. They're the one who makes it usable.
A note on the term. The paper uses “knowledge distance” for the gap between two human occupations — the dissimilarity between the skills their jobs demand. We use it for the gap between encoded business logic and the agent that must act on it. Different measurements; the same mechanism, and the same conclusion: past a certain distance, AI stops helping — and what remains is domain knowledge that cannot be substituted.
The readiness diagnostic reads your Knowledge Distance across all three dimensions — your people, your organization, your platform — and names where you actually stand. We're building a cohort of early adopters. If that's you, let's start the conversation.
Start the conversation →