The Organizational Commons is Adaptant's framework for adopting artificial intelligence in service of organizational intelligence, not of technology. First you design how decisions are made and who makes them; then, which technology amplifies them.
When electricity arrived, factories swapped the steam engine for an electric motor and kept the same layout of belts and pulleys. Decades went by without productivity gains, until someone redesigned the factory around what electricity made possible.
The same is happening with AI. Every area adds its assistant, every platform its meter, and the organization keeps working the same way: information travels upward, gets distorted at every hop, and decisions get made far from where reality happens. AI speeds up that journey; it doesn't change it.
And it makes producing cheap, but not understanding. Someone uses AI to turn three ideas into ten pages; someone else uses it to shrink them back to three ideas. More documents, more dashboards, more reports nobody opens: activity that registers as productivity and creates no value.
AI can concentrate power or distribute it. Technology doesn't decide that: organizational design does. That's why the Commons starts with design.
Intelligence isn't in the AI: it's in how the organization accesses, connects, and acts on its information. A company is intelligent when whoever faces the customer, the problem, or the opportunity has the information and the authority to decide. AI is the enabler of that autonomy, not an end in itself.
The framework doesn't start with the tool. It starts with four design decisions, made together with the teams that create value from the market.
The question isn't which AI tool to buy, but which decisions we want made closer to the customer, and what information they need to be made.
The framework rests on concrete infrastructure. It is not a dashboard nor a replacement for your systems. It's a layer that connects to what you already have — ERP, CRM, project management, finance — and consolidates it in one place where any member of the organization can ask in natural language.
"Commons Chat queries the Technology Commons Farm." — that's how the OC is spoken about externally.
A conversational datamart already exists — Copilot, Joule, Rovo. The Commons's difference isn't the interface. It's that it learns: from what people ask, from what they don't find, from what they correct, and from what the organization designs, delivers, and ships.
That capacity holds up because the Commons combines three sources that no traditional system integrates into a single living base:
The third source has its own mechanics: periphery capture — people tell the Commons what no system records, with human curation before perception becomes knowledge. It's one of the four loops through which the Commons learns: what gets asked, what is known, what is perceived, and what worked. The detail is in the framework.
The right question isn't "who do we give access to?" but "what kind of information requires what kind of care?" Position in the org chart shouldn't be the answer to either.
The destination is transparency: a mature organization operates with its information as universal access. Getting there is a learning process. That's why the Commons provides for opening cycles — differentiated access levels that exist as a transitional device while the organization matures its capacity to operate with open information. Any differentiation observed is a state of transition, not a permanent architecture.
| Type of information | Access | Criterion |
|---|---|---|
| Aggregate operational | Universal | Team results, area metrics, trends. The same context for everyone enables autonomous decisions and peer comparison. |
| Team operational | Team and peers | Internal metrics, project status, capacity. It opens toward universal as the organization matures its collective read. |
| Individual nominal | Direct relevance | Salaries, evaluations, identifiable data. In the opening cycle, salaries become visible first grouped, then by band, eventually nominal. |
| Sensitive strategic | Explicit agreement | Intellectual property, negotiations, data whose value is tied to its safeguarding. Economic and legal criterion — not hierarchical. The only level that doesn't migrate toward universal by design. |
Every platform you already use — SAP, Microsoft 365, Atlassian, ServiceNow — added its own AI consumption meter this past year, on top of the license you were already paying.
None use the same unit. None is easy to compare against the next. And in more than one case the vendor doesn't even publish the per-unit price until the invoice arrives. The industry already has a name for it: the "tokenization" of software cost.
The difference isn't a discount — it's structural. A single point of LLM consumption instead of one per system. Model abstraction: if a provider raises prices, you change one variable, you don't renegotiate under pressure. A floor near zero with a local model. A cache of frequent answers that the client controls.
The question for the CFO isn't "how much does the Commons cost?" but "how much are we paying today, summed across all platforms, for an intelligence that doesn't talk to itself?"
The Commons isn't installed and left running. It's an organism that goes through three stages: it's designed, it starts creating value, and it becomes self-sustaining in a team of its own. Each stage leaves the Commons more alive than the last — and in the end, the organization doesn't depend on Adaptant to maintain it.
The Commons Team. A living Commons needs someone to look after it. Not a new department or a layer of control: three roles that sustain access to information and the evolution of the AI, distributed among people who already know the domain.
Against conversational BI (Copilot, Joule, Rovo) — the difference isn't the interface, it's the combination of three things no vendor has together: the access-by-nature model, the integrated adoption methodology, and the active layer that learns. A BI answers inside its own system; the Commons cross-queries all of them.
Against each platform's native agents — the Commons doesn't compete on execution inside each system. It competes on consolidation: a single intelligence layer that cross-queries SAP, Jira, ServiceNow, and M365, instead of paying each one's AI meter separately. That's the FinOps conversation.
The Adaptant School trains executives, professionals, and consultants to design an AI adoption strategy in service of organizational intelligence. Over four cycles, participants work on why decentralized organizations perform better in complex environments, how to design a network of autonomous teams, and how to use AI as a tool for those teams. The Organizational Commons is the foundation of the second cycle. The program is taught in Spanish.