The last question is how a society would actually tell whether connection is working — and why the answer may decide the shape of the AI transition.
Measuring the network structurally
The wrong measure is the obvious one: number of followers, size of audience, volume of traffic. Those track reach, which the function has already shown is not the same as spread. The right measures are structural — how many participants can find relevant collaborators; how concentrated attention and network control are; how easy cross-domain connection is; how many disconnected clusters exist; how much capacity remains stranded; how portable identities and credentials are; how many routes connect local projects to wider resources; how dependent the ecology is on a few critical nodes; and how easily people can leave and reconnect. These are questions about the shape and health of the field, not the prominence of its stars — a network-specific instrument panel alongside the broader one ecological accounting builds, and subject to the same warning that a measure must never harden into a scoreboard.
Two measures are especially revealing. Propagation depth asks not how many people did this reach? but how many generations of new capability did it create? — teacher to student to new teacher to new students; founder to team to new founders to new projects — which is far closer to what ecological spread actually means. And the concentration-to-propagation ratio asks, when a project gains enormous attention, whether the benefit stays inside the original node financially, socially, and technically, or creates capacity elsewhere — the single clearest way to tell whether a system is amplifying a center or enriching the network. Recognition itself should propagate on the same logic: a healthy system helps valuable local work become visible to relevant networks rather than trapping recognition in elite institutions, which sets off the loop participation → recognition → connection → new opportunity → more participation.
What it changes about success
These measures quietly redefine success. Prevailing prestige asks how much did you personally accumulate? The model increasingly asks how much capable life became possible around you? This is not compulsory self-sacrifice — a great participant may still become powerful — but a different view of ecological effect, in which the deepest success is that a person’s existence makes many other capable participants more possible. It is the individual-scale reading of the whole function, and it is continuous with the model’s refusal to treat any participant as merely a node to be optimized.
How local capacity becomes civilizational
Connection is also what turns individual and local capability into the capability of a whole civilization. A child learns something, a teacher develops a method, a neighborhood tests a solution, a scientist makes a discovery, a project builds a tool — and if connection works, the capability travels, and others adapt, improve, combine, and teach it. That is how a civilization accumulates capacity over time, which makes this one of the main mechanisms of ecological growth. It is also how the self-steering ecology actually closes its loop: participants do not merely sense — a local observation becomes regional signal, national awareness, institutional correction — only if network pathways carry it. Connection is how distributed sensing becomes collective learning; without it, distributed intelligence stays isolated.
The state, and the many scales
Two structural notes keep the function from its own failure modes. As with aliveness, the state should support infrastructure, not own the network: government can legitimately back universal connectivity, interoperability standards, public communication infrastructure, anti-monopoly rules, open public knowledge, and portability protections — but it must not decide who everyone should connect with. The healthy shape is public support, distributed connection, minimal centralized control of content. And a healthy network needs multiple scales at once — local, regional, national, global; close relationships and weak ties; project, professional, and cultural networks — with no single scale dominant, so that a participant can live deeply locally while still reaching broader fields when they need to. Rootedness and reach are not opposites; the function exists to provide both.
The distribution question
All of this becomes decisive in the AI transition, which is why the function may matter more now than ever. If AI and robotics create enormous productive abundance, the question that determines everything is whether that capability spreads through the ecology or remains concentrated in the institutions that own the models, robots, data, compute, and platforms. The same technology that could distribute capacity more widely than any system in history could also centralize it more completely than any before — the architecture decides which. That is the gap the function names:
The difference between AI abundance and ecological abundance is largely a distribution-and-network question. AI can create enormous capacity; connection determines whether the ecology actually acquires it.
So the function is much larger than “networking.” Gathered into its formulations:
Aliveness creates living force. Connection gives that force pathways. Propagation turns local vitality into shared ecological capacity.
And structurally, the whole of it: the model should build an ecology where useful capacity can move from person to person, field to field, and generation to generation — without requiring central ownership, permanent dependency, or loss of differentiation. That is the architecture by which the ecology multiplies what becomes alive within it.