There is a problem with arguing that the powerful should care. That case still treats the existing power structure as the audience and asks how to persuade it — and plenty of people in power would rather be a big fish in a small pond: get rid of the agitators, keep the ones who submit. Ordinary authoritarian logic. Persuasion doesn’t reach it.

So the argument has to move down a layer, from morality to systems. The systems question is not how do we get incumbents to care? but: what arrangement produces the greatest long-run adaptive capacity of the whole, whether or not incumbent power prefers it?

Two optimizations that diverge

An authoritarian can be perfectly rational in preferring a smaller, more controllable pond, because their objective function is not the health of the ecology — it is the retention of control. That points at two different things a system can be optimized for:

  • Power optimization — maximize control, predictability, submission, and preservation of position.
  • Ecological optimization — maximize differentiated, viable, adaptive participation across the whole field.

These can pull hard against each other. A ruler often benefits from suppressing exactly the participants the ecology most needs — critics, inventors, dissidents, boundary-crossers, unusual thinkers, inconvenient truth-tellers. From the top, they look destabilizing. From the whole system’s view, they may be carrying essential corrective functions. Which makes the principle far more radical than “be generous”:

The ecology cannot afford to organize itself around the comfort of its dominant participants.

Their preferred equilibrium may be locally stable and globally degenerative at the same time.

What kind of friction is this?

A systems view doesn’t assume harmony is the goal. It can tolerate tension when the tension is carrying information the system needs. The agitator may be necessary; the dissenter may be carrying signal; the difficult child may be carrying an undeveloped capacity; the outsider may be exposing a legitimacy failure; the person refusing the arrangement may be registering that the form has drifted from reality.

So it never asks how do we eliminate friction? — the reflex criticized elsewhere. It asks what kind of friction is this?

  • destructive noise,
  • developmental stress,
  • a signal of missing legitimacy,
  • a structural warning,
  • merely competing jurisdiction,
  • or the system meeting a capacity it has no place for yet.

Telling those apart is nearly the whole art.

A sharper definition

That gives a better definition of the whole project than the moral one. Not everyone gets supported because everyone is valuable — too soft, too easily sentimental. Rather:

No ecologically significant capacity should be discarded merely because the current structure cannot comfortably accommodate it.

That is a systems principle, not a feeling. And it implies the system must actively preserve variation. Authoritarian structures simplify — fewer acceptable roles, narratives, centers of authority, tolerated behaviors — because simplification makes control easier. It also lowers adaptive capacity. A resilient system deliberately preserves heterogeneity, precisely because future conditions are unknown: the strange child, the obsessive researcher, the contrarian, the sideways thinker, the conventional stabilizer, the caretaker, the builder, the critic, the mystic, the systems thinker. You cannot know in advance which of them becomes essential under the next set of conditions.

Variation is not inefficiency. It is civilizational option value.

Seen this way, No Wasted Geometry sounds less like humanistic education and more like a theory of system resilience.

Power, redesigned

Which means power itself has to be redesigned — not eliminated, but distributed so that no single actor can permanently dominate the whole. Start from a blunt assumption: every concentration of power creates blind spots. It follows that healthy systems require competing forms of perception, jurisdiction, correction, and appeal. That is closer to ecology than to utopia. A forest does not work because one tree wins; it works because differentiated organisms fill different niches, exchange resources, constrain one another, and respond differently to shocks.

So the deepest move is to change what the whole thing is for:

A post-AI society is not built to maximize the welfare of individuals. It is built to preserve and develop the differentiated capacities through which the whole ecology stays intelligent and adaptable.

Individuals benefit enormously — but not as isolated units being optimized. And it quietly dissolves the big-fish objection. The big fish may prefer the small pond; the answer is simply that the pond is not theirs to define. Their preference is one input in the ecology, not its organizing principle. That is the boundary between a power-first logic and a systems logic.