Kodak: The Origin

In two thousand four I built my first server. It was not in an air-conditioned office with a corporate budget: it was a secondhand computer, with CentOS installed by hand, inside a Kodak photo lab in Matamoros, Tamaulipas. I processed digital images at volume — five hundred, six hundred eight-by-ten prints a day, thousands of five-by-sevens for schools and photographers from across the region. And every day I watched the same bottleneck, so absurd that no one questioned it anymore: photographers lost two and three whole days traveling to deliver their physical memory cards, or shipped them by courier and prayed they would arrive.

I thought about the problem like an engineer, not like a resigned photographer. If the file is digital, why does it travel as if it were an object? I stood up an FTP server with a user account for each photographer, domain management, storage spaces. From that night on, they uploaded their work over the Internet at dawn, the lab printed the next day, and the photos went out by courier already developed. I eliminated, in one stroke, two or three days of logistics everyone took for inevitable. I invented no new technology: I took tools that already existed and put them at the service of a small business no one considered worthy of them.

That was the moment I understood something that would become the axis of my whole professional life, and much later, of this book: enterprise technology can change the lives of small businesses, but no one offers it to them, because «they aren't corporations». The big providers look upward, toward the accounts that bill millions. Below remain thousands of photographers, accountants, farmers and trades who would do wonders with half that infrastructure, if someone bothered to translate it for them.

But there is a second lesson in the Matamoros lab, and it is the one that truly ties this chapter to the rest of the book. That server was not a promise of what the lab would someday become. It was, already, the proof of what the lab was: a place that had decided to solve the real problem of its people. The act defined the identity at the very moment of its execution. Two decades before I could name it, I was already living the Contrarian Thesis with my hands full of cables. Philosophy did not reach me from books. It reached me from acts, and the books only came afterward to confirm what the hands already knew.

Francisco in Huntsville

I met Francisco Hernández at an event, through a mutual friend. He was over fifty, a photographer in Huntsville, Alabama, and he carried a problem he himself did not see as a problem: his workflow was chaos, and he had grown used to chaos. He stored the SD cards of each wedding in physical envelopes — one envelope per client, with several cards inside: camera one, camera two, the drone— and stacked them in a shoebox. Every time he had to edit, he opened the box, looked for the right envelope, put the cards into adapters and plugged them into the computer one by one. He did not know which cards were full and which empty. He worked with discontinued software, already incompatible with his system, and still did not dare to change, because «that was how he had always worked».

Francisco did not come to me because he had lost data. He came because he wanted to learn to edit better, and he took a week-long in-person course I taught — Premiere, Photoshop and, above all, workflow. What truly convinced him was not a slide or a promise: it was seeing the server working live. That is a constant in my craft: people do not buy what they do not understand, and they do not understand infrastructure until they see it breathe. He saw two computers editing at the same time, without passing files over USB, all reading from the same place at ten gigabits per second. And he understood, in the body, what no technical explanation had conveyed.

I built him a Dell PowerEdge R five ten server, dual processor, twenty-four terabytes in a RAID five array of eight four-terabyte SAS Enterprise spinning drives configured with redundancy, and a ten-gigabit network with a fiber switch. I flew from Seattle to Atlanta and drove four hours to Huntsville to install it in a single day. The total cost — server, switch and installation— came to around two thousand seven hundred dollars. After one year, Francisco left the shoebox behind. Today, seven years later, he still uses the same server, with two editing stations and two client photo-review stations working simultaneously, and I give him remote maintenance now and then. His words, which I do not forget, were: «I don't know how I got so far with those envelopes and memory cards. Now I work on several projects at once and I don't worry about anything.»

Francisco's story is the chapter of habit made flesh. For years, his identity as a photographer was defined by an inherited habit — the shoebox— that he had never consciously chosen and that, nonetheless, revealed him: he was an artist trapped in an amateur's logistics. The day he changed the habit, he changed who he was in his craft. He did not become another photographer in the future; he was another photographer from the first edit on the new server. The habit did not promise his destiny. It revealed, and then remade, who he was. That is the entire thesis, told with a Dell PowerEdge R five ten and an empty shoebox.

The Sovereignty of Data

If Francisco's story is habit at human scale, the one that follows is the same truth at company scale, with a bill that hurts. An agricultural-technology company I work with — from the world of apples, here in central Washington, where I live— had its services scattered across the cloud and paid around nineteen hundred dollars a month. Then its cloud provider offered what looks like a gift and is almost always a hook: a free year. They accepted, as anyone would. They migrated virtual machines, storage, processing. And when the free year ended, reality's projected invoice arrived: twenty-two to twenty-four thousand dollars a month.

That jump is no accident; it is the business model. The cloud hooks you cheap when you are small, lets you grow comfortable, and charges you the fortune precisely when leaving costs more than staying. When you no longer own anything — not the server, not the system, not the place where your data lives— the price stops being negotiated with you and begins to be imposed on you. You do not pay for what you use; you pay for not being able to leave.

I sold them twelve Dell servers and built their complete infrastructure, on-premise: email, databases, internal applications, storage. The return on investment came in three or four months. Against a bill of twenty-four thousand a month, the annual savings are counted in hundreds of thousands of dollars. But the number, brutal as it is, is not the important thing. The important thing is what they recovered along with the money: total control over their own data — sensitive data, agricultural, client data— and the freedom not to depend on conditions a third party might change tomorrow without warning.

Here the thesis leaps from the existential to the technical without changing shape, because it is the same principle. I claim that your identity is what you control and practice today, not what you promise to be tomorrow. Well then: to own the infrastructure your data runs on is to own the evidence of who you are; to rent it from a third party is to surrender the record of your identity. The cloud is a cheap hook that ends up costing a fortune; data sovereignty means tangible ownership of your information, on infrastructure you control. You own, you do not rent. It is the exact structure of the motto, translated into iron and optical fiber: what you own and what you practice define who you are — not what someone promises you, nor rents to you at rationed prices, for a tomorrow that never quite arrives.