The System of Work

I am developing a model for the conditions that help people do meaningful things together. I call it the System of Work. It connects software, organizations, culture, leadership, and everyday coordination because, in practice, a problem often crosses all of them.

The purpose is human: more capacity to act, create, belong, and learn together. The model is my current explanation of what helps that happen. It is being shaped through building and operating real things, comparing experiences, teaching, and finding places where the explanation fails.

This page is an orientation to that work as of September 2026. The detailed arguments and cases carry their own context and uncertainty.

Start with the situation

Consider a person who keeps holding a team together by remembering every exception, answering every question, and making sure nothing falls between responsibilities. They may be doing excellent work. Their repeated intervention also tells us something about the system around them.

Perhaps people cannot find the context they need. Perhaps responsibility is visible but authority is missing. Perhaps the tooling makes the right action expensive. Perhaps the group has avoided a conversation that no tool can resolve. Those explanations imply different interventions. Calling all of them a productivity problem would hide the useful distinctions.

The first move is to understand the actual situation: who is trying to do what, what keeps getting in the way, who absorbs the consequences, and what participants already do that works. A familiar pattern is a place to look. It still has to earn its relevance here.

What the model pays attention to

Several questions keep proving useful across my work:

• Intent: What are we trying to make possible, and what matters when trade-offs arise?
• Context and memory: What does someone need to understand now, and what must remain available later?
• Responsibility and authority: Who can act, who is accountable, and where does a decision need help?
• Trust and care: Can people ask, disagree, recover from mistakes, and rely on one another without giving up their agency?
• Coordination: How do separate actions become a coherent result, including when circumstances change?
• Tools and environments: What do our defaults, interfaces, spaces, and rituals make easy or difficult?
• Learning: What happened, why do we think it happened, and what should change as a result?

These questions describe areas to investigate. They are not a claim that every system has the same components, or that one structure will work everywhere. Some tensions, such as autonomy and shared responsibility, stay with us. Good systems make them understandable and navigable.

How the practice develops

My working method is to connect observation and action in a continuous loop. Clarify the problem, form a provisional explanation, make a useful change, observe its effects, and revise. Preserve enough reasoning that the next person, or my future self, can understand the decision without reconstructing the whole history.

An intervention should provide value at its current scale. It might be a small software capability, a better handoff, a boundary around a role, a conversation, or a change to a physical environment. The larger direction keeps these increments coherent; it does not excuse postponing usefulness until a grand system is finished.

I describe this method in Building Through Emergence. Adaptive Collaboration is a situated investigation growing out of Titanic's End: what happens when social attention keeps moving, while tools ask people to maintain fixed groups and separate stores of information? Its proposed model remains something to test, and the community's behavior can change the answer.

Where humans and AI enter

AI can participate in work, produce artifacts, recover context, and carry out increasingly substantial actions. That makes clear intent, boundaries, inspectable state, and ways to notice mistakes especially consequential. A faster participant can amplify both a useful system and a confused one.

There are recurring coordination questions across human and AI participation, but their differences matter. They have different capabilities, failure modes, needs, and responsibilities. A shared operating context does not make them interchangeable. The system must serve the people whose lives and work it affects.

I am interested in what this makes possible beyond faster execution: less attention spent recovering lost context, more room for judgment, more ability to contribute, and better ways to learn together. Those are desired outcomes. They need evidence from use.

How to read and challenge the work

The ideas here come from multiple settings. Software and platform work, volunteer operations, community life, and coaching can illuminate one another, but a lesson does not become universal by appearing in a compelling story. Cases should show the situation, my contribution, other people's contributions, the decisions, what was observed, and what remains uncertain.

A proposal demonstrates an intended intervention. Code demonstrates a mechanism. Neither alone demonstrates adoption, reduced burnout, or a healthier culture. I want those distinctions to stay visible as the body of work grows.

The most useful response is often a specific counterexample: where a pattern failed, what conditions changed it, or what a simpler explanation captures. New evidence can change a local practice or the underlying model. Keeping that history makes the work more useful than silently replacing an old conclusion.

A working test in this library

This library is one small application of the model. Private observations can keep developing while a public paper or a manuscript preserves the revision its reader received. One shared body of work does not require every view to expose the same material or advance at the same time. The useful boundary is explicit: preserve identity and provenance, then decide which interpretation and revision belong in each setting. Whether this makes ongoing thinking and publishing easier remains something to learn through use.

Over time, the connected material can support a book, deeper guides, or teaching. Their value will be in helping someone build judgment in their own context. The work succeeds when people can use it, question it, and develop ideas I would not have reached alone.