Once people accept that fragmented knowledge is a structural problem, and once they see the value of a shared semantic core with tailored experiences, the next question is usually practical.

Where do you actually start?

This is the point where a lot of good thinking collapses into either overreach or caution.

Some organisations respond by imagining a total enterprise redesign. They picture a complete canonical model, enterprise-wide agreement on terminology, full architecture alignment, and a new contribution system for everyone all at once.

Other organisations respond by making the idea so modest that it becomes harmless. They run workshops, produce conceptual diagrams, talk about knowledge reuse, and maybe create a pilot artifact or two — but without changing how contribution actually accumulates.

Neither response is strong enough.

If organisations need a shared model, there is an obvious danger waiting right behind that insight.

People hear “shared model” and imagine a single master interface, a single canonical workflow, or a single view of the organisation that everyone is somehow expected to use.

That is not the answer. In fact, it is one of the fastest ways to make a shared model fail.

The right pattern is not one generic experience for everyone. It is a shared core with tailored experiences.

If the first problem is fragmentation, the next question is obvious: what actually changes when an organisation stops treating collaboration as the exchange of disconnected artifacts and starts treating it as contribution into a shared model?

A lot changes.

Not because the organisation suddenly becomes perfectly aligned. Not because everyone adopts one tool. Not because all ambiguity disappears.

But because meaning begins to accumulate instead of being repeatedly recreated.

Most organisations do not have an information shortage. They have a coherence shortage.

Across the enterprise, teams are constantly producing knowledge about the same underlying reality: customers, services, products, systems, work, risks, controls, priorities, dependencies, and performance. Yet this knowledge rarely accumulates as a shared organisational asset. Instead, it is created in local tools, shaped for local purposes, and maintained under local pressure.

The result is familiar. Different teams hold overlapping versions of the truth. Each version is useful within its own context, but weakly reusable outside it. Business teams describe capability one way. Service teams describe it another. Architects model it differently again. Operations teams report on outcomes through yet another lens. Everyone is working from information that is materially related, but structurally isolated.

This is one of the defining operating conditions of modern organisations: not ignorance, but fragmentation.

A good way to describe what changed is this:

AI is starting to feel like a Star Trek replicator for ideas.

Not in the sense that it makes reality appear instantly. Not in the sense that execution stops mattering. Not in the sense that every output is automatically good.

But in the sense that a large part of the translation work between idea and actionable form just became dramatically cheaper.

That is a very big deal.

The AI conversation is still too often trapped in procurement language.

Which vendor. Which model. Which platform. Which copilot. Which controls.

Those questions matter. But they are no longer the main event.

The real break is this:

For the first time, organisations can start building meaningful new capability from within, with far less effort than before.

That should completely change the ambition.

One of the biggest changes AI introduces is not just automation.

It is attention.

For the first time, organisations can start operating with something closer to persistent internal attention at scale.

Not perfect judgement. Not independent wisdom. Not magic.

But real, ongoing attention that can keep looking, connecting, drafting, checking, refining, and improving.

That is a much bigger shift than most AI strategies currently admit. And it is still badly under-theorised in most organisations.

When people hear the phrase organisational gardens, it can sound soft, vague, or decorative.

It is not.

It is a practical way of talking about the parts of an organisation that only stay healthy when they are continuously tended.

That is a large part of what makes an organisation actually work. And it is exactly the terrain most organisations chronically under-invest in.

Most organisations do not have a technology problem when it comes to AI.

They have a context problem.

The knowledge of how the organisation really works is scattered across people, systems, habits, exceptions, workarounds, and fragments of documentation. Important decisions are often poorly connected to the workflows they affect. Teams compensate socially. They ask around. They rely on memory. They improvise.

That is survivable in a human-only organisation. It becomes much more visible when AI enters the picture.

AI quickly exposes whether the organisation has a coherent enough internal structure to support shared understanding, guided action, and repeatable improvement.

That is why more organisations will need to build something like a KnowledgeFund for themselves.

A lot of organisations are still treating AI capability as if it were something they can procure.

Buy the platform. Sign the enterprise deal. Enable the seats. Approve the use cases. Run the rollout.

Then call that capability.

That logic is already breaking.

The models matter. The infrastructure matters. The vendors matter.

But the real advantage is moving somewhere else.

You do not mainly need to buy the capability. You need to build it.

That is the shift many AI strategies are still refusing to admit.