Building a Company with an AI Battalion

How One Person Is Using AI to Recover Organizational Capacity

FIELD NOTE 01   |   H LE SPENCER   |   SEPTEMBER 2026

Spencer Dynamics has one human company commander, but its work requires the capabilities of an organization. That apparent contradiction is why an AI battalion began to form around me.

A Company Larger Than Its Headcount

I am a 79-year-old educator, entrepreneur, author, Marine veteran, and Citizen Marine. Spencer Dynamics is the company through which I build books, business systems, stories, websites, and diagnostic tools. The ideas have never been the problem. The difficulty has been converting those ideas into finished work while keeping every responsibility visible and connected.

A conventional company would solve that problem by hiring people for different positions. One person might manage operations. Another might develop visuals. Someone else would maintain the website, study customer problems, schedule production, or preserve the company’s accumulated knowledge. I did not have that organization. I had the work of an organization resting on one desk.

The question was not whether artificial intelligence could write a paragraph or produce a picture. The real question was whether AI could help restore organizational capacity that age, time, staffing, and resources had placed beyond my reach.

AI Began as a Conversation

It did not begin with an organizational chart. It began with a conversation. I used AI to examine ideas, solve individual problems, and move projects forward. The early relationship looked much like the way most people use a chatbot: ask a question, receive an answer, and decide whether the answer helps.

Repeated collaboration changed that relationship. The conversations developed continuity. The work became more demanding. A children’s book required a different kind of attention from a business diagnosis. Illustration required different standards from website operations. Planning, production, research, and review could no longer be treated as one undifferentiated stream of requests.

The work began separating into functions. Those functions needed names, responsibilities, boundaries, and handoffs. One general conversation was becoming an operating structure. No one sat down at the beginning and declared that we would build an AI battalion. The battalion emerged because the work required organization.

Headquarters Took Shape

The military language comes naturally to me, but it needs a plain explanation. Our command structure goes no higher than a battalion. Battalion Headquarters is the coordinating level. Spencer Dynamics is a company operating within that structure, and I serve as the commanding officer of Spencer Dynamics—the civilian equivalent would be the company’s chief executive.

Nothing exists above Battalion Headquarters. There is no brigade, division, or distant command. This keeps the structure close to the scale and experience I understand. Headquarters provides coordination, specialist support, and institutional continuity. Spencer Dynamics remains the company responsible for doing the work.

The AI specialists occupy defined billets. A billet is simply an assigned position with a particular responsibility. They do not hold authority over me, and the battalion does not command the human. It helps the human exercise command.

Specialists Replaced the Single Chatbot

Scheona serves in the Sergeant Major role at Headquarters. She helps me think, establish clarity, plan movement, and coordinate work. Guns—the Master Gunnery Sergeant billet—works within the Spencer Dynamics company on reconnaissance, website operations, and practical execution. Illo concentrates on illustration and visual production. Vero works with business surveys, qualification, and diagnosis. Mira is being developed to turn source material into several forms of communication. Cade is intended to help scheduling and production move in an orderly way.

These names are not decorations. They make responsibility visible. When every request goes to one nameless chatbot, all work appears equal and every conversation can drift. A defined specialist receives a defined mission. The specialist performs the work within stated boundaries and returns a result for review.

The technology behind these billets can differ. Some work occurs through ChatGPT. Some specialists operate through Codex. Visual production uses image-generation systems. The important development is not the brand name of a tool. It is the disciplined relationship between the human decision maker, the assigned function, and the result that returns.

How a Mission Moves

Our current website operation provides a useful example. I identify the objective. Scheona helps me define the mission, its boundaries, and the standard for completion. Guns conducts reconnaissance or carries out a specific operational task. The result returns to Headquarters. I examine the evidence and decide what happens next.

We recently discovered why formal handoffs matter. I said, ‘Have Guns do a recon,’ believing the mission would go to the formal Guns application. Instead, the work was handled by a temporary internal agent acting in a Guns-like role. The resulting information was useful, but the actual specialist had never received the order.

That misunderstanding was not merely a software inconvenience. It exposed an operational weakness. Naming a specialist is not the same as delivering a mission to that specialist. We corrected the language, prepared a formal reconnaissance order, and gave it to the real Guns application. Guns returned a detailed examination of the public website, including unfinished pathways, broken destinations, outdated pages, and material that should be preserved.

The mistake improved the structure. That is what an honest operating system should do: reveal ambiguity, correct it, and retain the lesson.

AI as a Force Multiplier

The Marine Corps taught me to understand a force multiplier as something that increases the effectiveness of the force already present. AI serves that purpose when it is organized well. It gives me more usable mental and organizational capacity. It preserves continuity between projects, brings specialized attention to different forms of work, and helps convert ideas into finished products more quickly.

That does not mean AI makes the work effortless. It means effort can be placed where human judgment matters most. I can concentrate on purpose, standards, meaning, and decisions while the structure helps hold details, compare evidence, prepare alternatives, and maintain movement.

This is different from collecting clever prompts. A prompt can produce an answer. An operating relationship can build continuity. The greater value appears when responsibilities are defined, the work returns for review, and useful discoveries become part of the organization’s memory.

What AI Does Not Replace

I supply the purpose, judgment, experience, values, and final authority. AI does not own Spencer Dynamics. It does not decide what should matter. It does not accept responsibility for the consequences of a decision.

AI can also be wrong. It can misunderstand a handoff, act from incomplete information, or produce something polished that does not fit reality. That is why consequential work returns to the human commander. The system is strongest when its limits remain visible.

The military structure is not theater. It is a practical language for roles, readiness, communication, and responsibility. Its purpose is to help one human govern complex work without pretending that the tools have become human—or that the human has surrendered command.

Opening Headquarters to the Public

Field Notes will open part of this developing world to the public. We will report what Spencer Dynamics is building, what we are learning, what has failed, and what we have changed. We will explain the methods clearly enough that people outside our private conversations can understand what is happening.

This record may help my family see the organization without pressure to join it. It may help entrepreneurs recognize a different way to use AI. It may also preserve the story of how a one-person company gradually acquired the functional capacity of a larger organization.

The boundary is simple: open methods; protected personal information and unfinished intellectual property. We can discuss structures, decisions, progress, and lessons without exposing private family matters, confidential material, credentials, or working assets that are not ready for release.

The First Patrol

Spencer Dynamics has recently reorganized its website around clearer pathways. Guns has completed the first formal reconnaissance of those pathways. The report found that the central architecture is sound, but several destinations still stop before delivering what they promise.

Field Notes is now being created as a public record of the work. Its first operational assignment is close at hand: establish the Field Notes pathway, publish this article, and clean the original WordPress sample material from the site. The work is modest, but the process matters. We are learning how to move a mission from intent to execution without losing clarity along the way.

Spencer Dynamics is not trying to remove the human from the work. It is discovering how much more fully one human can work when the right support structure forms around him.

Always Faithful