From Scribes to Swarms: AI Is Our Printing Press

Before the printing press, knowledge moved at the speed of a hand. Scribes sat with ink and those big feathery pens, copying page after page so a book could exist twice. It was skilled work. It was slow. For people who loved letterforms, layout, and the quiet discipline of the desk, it was identity as much as employment.

Then Gutenberg’s press arrived. A machine could multiply text faster than any workshop of pens. Demand for hand-copied manuscripts collapsed. Livelihoods changed. The artisanal spirit of the scribe did not vanish overnight, but the economic center of gravity moved. Reproduction stopped being the bottleneck. Ideas could travel.

That story is not nostalgia. It is a map for what is happening to knowledge work right now.

I Have Always Been Drawn to How Things Work

I am an engineer by temperament. Long before titles and companies, I wanted to know what made a system tick: the gears, the loads, the failure modes, the path from intention to result. In college I studied mechanical engineering and graduated with that degree. Along the way I kept drifting toward computer science.

The attraction was not that software was “easier.” It was speed. In information technology you can invent something in the morning, ship a rough version by afternoon, and learn from reality before the week is out. That feedback loop is intoxicating if you care about building. Hardware teaches patience and physics. Software teaches iteration. I wanted both, and I have spent a career across platforms, data, security, operations, and the human systems that sit on top of the tech.

I have also lived through two previous “this changes everything” waves firsthand: the internet, then cloud computing. Both were enormous. Both rewrote industries.

I believe the AI transformation is bigger than either of those, and bigger than the printing press analogy alone can hold. It is not only a new distribution channel for information. It is a new way to produce work.

The Scribe Moment for Knowledge Work

If the printing press reduced the need for scribes who copied text, AI reduces the need for humans whose primary value is copying process: drafting the first version, summarizing the pile, stitching the ticket, chasing the status, rewriting the same policy for the twentieth customer.

That is disruptive for anyone whose identity is tied to “doing the writing” or “running the checklist” by hand. It can feel like a loss of artistry. I understand that feeling. Craft, taste, and judgment still matter. What changes is where scarce human attention should go.

The press did not eliminate authors. It eliminated the bottleneck of reproduction so authors and readers could meet at scale. AI will not eliminate leaders, engineers, operators, or builders who understand how systems work. It will eliminate the excuse that we cannot afford to try, document, coordinate, and improve at the pace the market now demands.

My AI Brain and the Swarm

I did not want a single clever chatbot in a browser tab. I wanted an operating system for attention: a shared brain, a swarm of agents with clear jobs, and a management model that keeps the whole thing honest.

A shared knowledge vault. This is the company brain: decisions, project context, playbooks, loops, and living memory. Agents do not start from zero every morning. They read from the same source of truth and write back so the next shift inherits the work.

Specialized agents, not one generalist. The swarm is organized by purpose. One agent acts as chief of staff: routing work, watching the task board, reporting status, and keeping coordination from turning into chaos. Another owns engineering across the products I care about. Another owns day-to-day operations for marine and marketplace businesses. Another owns intelligence and monitoring: traffic, visibility, health checks, and overnight drift signals. Different machines, different toolkits, one roster.

Management is explicit. Agents do not “just figure it out.” Work lands on a shared task board. Recurring jobs run as loops with heartbeats. Attention files flag when something needs a human. I stay in the loop as the principal: priorities, judgment calls, and what is good enough to ship. The swarm multiplies throughput. It does not replace ownership.

Shared memory, role clarity, and lightweight governance turn AI from novelty into leverage. That is also where most organizations stall. They buy tools. They skip the operating model.

What This Means If You Are Leading a Company

The question is not whether AI will rewrite how your team works. It already is. The question is whether you will design the transition or inherit it.

Design means clarifying artisanal judgment versus reproduction. Building a knowledge layer people and agents can trust. Pairing AI with real operating cadence: security, quality, cost, and customer outcomes, not demos. Leading with both truths at once: honor the craft, and refuse to pretend the pen is still the bottleneck.

If your organization is staring at that shift and wants a partner who has lived internet, cloud, and now AI from the operator’s chair, that is the work I do through Advisory Services. Depending on depth of need, that can mean Advisory CTO counsel with the CEO or C-suite, Fractional CTO partnership alongside your team, or Interim CTO coverage when you need full-time leadership in the seat.

The printing press did not ask the scribes for permission. AI will not ask us either. The opportunity is to keep the spirit of the craft (curiosity about how things work, pride in quality, care for the customer) while building the new presses our companies actually need.

I am building mine in the open, one agent and one loop at a time. If you want help building yours, start a conversation.