Projects are a good test case, because they expose the difference between organizing around roles and ownership and organizing around capacity, fit, and contribution.
Today, most projects begin with the project itself: what needs to be built, who owns it, what’s the budget, what jobs do we need, who has the credentials, who reports to whom, what’s the deadline? Then people are fitted into that structure. Under No Wasted Geometry a project still needs goals, deadlines, accountability, budgets, and leadership — but the design logic starts elsewhere: what is trying to be accomplished, what functions does that require, and which participants can carry those functions without needlessly wasting their other capacities?
The two models
| Today’s project model | No Wasted Geometry project model |
|---|---|
| Decomposed into jobs | Decomposed into functions and capacities required |
| People apply for predefined roles | Participants matched where they can contribute coherently |
| Credentials act as major filters | Demonstrated capacity, potential, and fit matter more |
| One person usually occupies one role | A participant may carry several functions, or shift as the project evolves |
| Employer owns the work structure | Ownership and decision jurisdiction can be separated |
| Manager distributes authority | Authority follows relevant competence and legitimate jurisdiction |
| Salary largely defines value | Compensation supports participation but doesn’t fully define contribution |
| People hide weaknesses | Limits are useful information for project design |
| Difficult personalities are removed | First ask whether friction signals bad fit, missing support, conflicting jurisdiction, or a real problem |
| Training prepares people for the assigned job | Formation helps participants become more capable contributors |
| Success = delivery on time and budget | Success = delivery plus increased ecological capacity |
| After completion, the team dissolves | Developed capacity, relationships, and knowledge stay in the ecology |
A worked example: a transit system
Today a government or company commissions it; consultants design, engineers engineer, finance controls the money, contractors build, community engagement happens at designated points, and everyone stays within professional categories.
Under No Wasted Geometry you keep the objective — a reliable transit system — but map the required ecology first. You need people who can carry structural viability, resource judgment, technical execution, community legitimacy, real knowledge of riders, long-range planning, environmental sensing, system integration, conflict resolution, and operational maintenance. None of those are necessarily job titles. A resident who has ridden the bus for twenty years may hold extraordinary system knowledge without a transportation credential. A brilliant structural engineer may be terrible at judging community legitimacy. A politician may have jurisdiction to authorize funding without jurisdiction over technical design. A firm may own the construction machinery without thereby earning the right to decide which neighborhoods get displaced. Which is the crucial principle:
Contribution, expertise, ownership, and decision authority are related but not identical.
Today they’re usually bundled. No Wasted Geometry unbundles them — the same move it makes with ownership generally.
Leadership becomes functional
This does not abolish leadership — it demands clearer leadership, because jurisdiction has to be explicit. Instead of “I’m the CEO, so the final call is mine,” you get: who owns the integration decision? who has technical jurisdiction? who represents affected participants? who judges resource viability? who can halt the project for safety? Different people may lead different dimensions, and there is still an integration function — someone, or some body, able to make the final whole-system call when functions conflict. That’s not egalitarian consensus; it’s differentiated governance.
Joining, failure, compensation, ownership
- Joining. Instead of only job → résumé → screening → employment, a person carries an evolving capacity profile (competencies, interests, developmental needs, previous contribution, working conditions, limits, learning edges), and projects publish functional needs rather than only jobs — “we need someone who can translate technical complexity into language neighborhood groups can actually use,” or “someone who can find where our assumptions fail under real operating conditions” (the strong critical thinker today labeled “difficult” gets a legitimate place).
- Failure. Before can they meet the role or should we replace them? come earlier questions: is the assignment wrong? is a resource missing? are we asking one person to do a different function’s work? are they capable but not yet formed for it? is jurisdiction unclear? is the environment suppressing the very capacity we recruited? Sometimes the answer is still they can’t do this and should leave the role — that’s allowed. The difference is that role failure doesn’t automatically become participant failure. You preserve the geometry while correcting the placement.
- Compensation. A project still consumes time, energy, expertise, and risk, so pay remains appropriate — but it becomes payment for specific contribution and cost, not the source of someone’s basic right to exist in the ecology. If basic viability is secured independently, compensation gets more honest — what is this contribution worth? rather than how little can we pay someone who needs this to survive? People can refuse bad projects more easily, so bad projects struggle to attract capacity. A useful feedback loop.
- Ownership. Capital is a real contribution and deserves standing and return — but so do technical invention, the workers, the community absorbing externalities, future users, public infrastructure, and environmental constraints. Capital becomes one participating function, not the sovereign function, which points toward mixed structures: private equity plus worker stakes, community stakes, public claims, contributor royalties, shared IP.
AI: which capacity can it release?
Today organizations ask which jobs can AI eliminate? An NWG project asks which human capacity can AI release? If AI handles the tedious documentation, the engineer spends more time on hard structural problems; if it handles routine coding, the programmer works at the architecture level; if robots take the dangerous construction, people concentrate on judgment, maintenance, and invention; if it lets a teenager with no credentials model something sophisticated, previously invisible capacity enters the project. So AI is measured not by headcount reduced but by additional ecological capacity made available — the same fork as reveal, not judge.
The project-level principle
The clearest summary of the difference:
- Today’s project: assemble enough labor and capital to produce an outcome.
- NWG project: assemble the right differentiated capacities to produce the outcome while leaving the participating ecology more capable than it was before.
That second clause carries the weight. A project that ships a brilliant product while burning out everyone, hoarding all the knowledge in one firm, exhausting the community, and leaving participants less capable is a success by today’s measures — and a poorly designed project by this one, because the product succeeded while the ecology lost capacity. Which gives a strong, portable test:
A good project produces its intended result without unnecessarily wasting the geometry required to create it — and ideally leaves the surrounding ecology more capable than it found it.
And once that becomes the goal, it can become a selection method: if we can measure ecology, we can predict which projects will leave the field most capable and choose those. That is worked out in Choosing Projects by Ecological Return.