> ## Content Index
> Fetch the complete content index at: https://data-decisions-and-clinics.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Where the Time Came From
- URL: https://data-decisions-and-clinics.com/where-the-time-came-from/
- Published: 2026-08-12T16:33:30.000Z
- Updated: 2026-08-12T16:34:14.000Z
- Description: People keep asking how I found time to write four books in a year while doing my day job. The answer is not a tool and not discipline. It is a layered setup where every piece of work is tagged, linked, and ready for the next one. The tools are rented. The brain underneath is the asset.
- Author: Yoram Friedman, MD
- Tags: Product Management, people

A question I get often recently: how did you get the time to write four books in the last year, on top of the day job?

The answer people expect is either a tool or a superpower. The honest answer is neither. The time came from a structure, and the structure is available to anyone willing to build it. This piece is my attempt to describe it plainly, because I keep watching smart product managers, analysts, and domain experts buy the same tools I use, get mediocre results, and conclude the technology is overhyped. The technology is fine. They stopped building one layer too early.

## The floor everyone is standing on

Start with what the tools alone are worth. Every AI assistant on the market will summarize your meeting, draft your email, restate your document. This is real value and it is worth nothing competitively, because it arrives for everyone on the same day. A spreadsheet made every analyst faster and made no company better at analysis than its competitors. The gain was universal, so it became the floor. AI chat is this decade's spreadsheet: if your entire setup is a subscription and a chat window, you have reached the floor and you are standing on it with everyone else.

That chat window is what I call layer zero, and most people never leave it. They set up the tool, ask it questions, get generic answers in a generic voice with no memory of their work, and conclude the tool is mediocre. It is. At layer zero, mediocre is the ceiling, because the model knows nothing about you except the last message you typed. I think of this as the mediocrity plateau, and the tell is a sentence I hear weekly: I tried it, it was fine, it saves me a few minutes on emails.

## The layers above the plateau

What sits above the plateau is a stack of layers, each one an investment, each one compounding the ones below.

The first layer up is identity and voice: teaching the system who you are, how you write, what your role demands, what good looks like in your world. A few hours of deliberate work, refined over months of corrections. From that point the output starts sounding like you on the first draft instead of the fifth, and every downstream layer inherits the gain.

Above the brain, which deserves its own section, the layers are cheaper and faster: connections that keep everything current instead of stale, and repeatable procedures written once so recurring work is done the same way every time. At the top sits the working loop itself, the part I have described elsewhere as an active-learning pattern: multiple systems in distinct roles, research triangulated across instruments, one model drafting, a different model validating, and me as the supervisor making the only calls that matter, what is true, what is mine, what ships.

![](https://storage.ghost.io/c/e6/05/e605b234-3cc8-4fe4-97bb-2bb55b18b8fb/content/images/2026/08/Untitled.jpg)

## The brain is the asset

The second layer is the one where the real value lives, and it deserves its own name: the brain. This is your body of work, organized so a machine can actually use it. On my authoring side, that means every article I have written cataloged with its core argument and evidence, every framework named in a controlled vocabulary, every book chapter cross-linked to the articles that seeded it, all of it connected with deliberate ontology and semantics. On my work side, it means the documents, the tickets, the wikis, the internal policies, the product standards, the meeting summaries, connected so that the assistant reads what is true this week instead of what was true when a model was trained. The better the brain, the better every single interaction on top of it. A modest model with my catalog beats a frontier model without it for any question that touches my own work, and most of the questions that matter touch my own work.

To make that concrete: my authoring brain currently holds every article I have published, a catalog of more than 170 entries where each piece carries its core argument, its key evidence, and the frameworks it uses, a controlled vocabulary of 60 named frameworks so no concept gets two names, and four books cross-linked to the articles that seeded them. None of this is exotic. It is markdown files in folders, maintained by one unglamorous habit: the day a piece publishes, it gets its catalog entry, twenty minutes of work while the piece is still fresh in my head.

![](https://storage.ghost.io/c/e6/05/e605b234-3cc8-4fe4-97bb-2bb55b18b8fb/content/images/2026/08/Untitled1.jpg)

That habit pays out in ways that surprised me. Before a difficult meeting, I search the catalog for the framework or the dated study that fits the moment, and it returns answers I had forgotten I had given. When I start something new, the first question is what the catalog already holds on the topic, and the second is what it doesn't; a topic with no entry is sometimes the most useful thing the catalog can tell me, because the gap is the next piece.

## It takes structure to make speed

Here is why this answers the four-books question. There is an old observation in sociology called the Matthew Effect: advantage accumulates to those who already have it. It is usually told about money, and the folk version is that it takes money to make money. In this system it takes structure to make speed. Every article I write is cataloged the day it publishes, which means every article is raw material for the next one. A piece on evaluation seeds a chapter on testing. A chapter on supervision surfaces a gap the catalog shows I have never covered, and the gap becomes the next article. Nothing is ever written from zero, because everything I have already written is tagged, linked, and sitting where the next piece of work can consume it. The books did not take time away from the articles. The articles compounded into books. That compounding is not available at layer zero, at any price, with any tool.

## The layers decay, and the tools do not matter

Two honest warnings belong in this picture, because the structure is not free and it is not permanent. The first is that the layers decay. Voice profiles drift as the models underneath them change. Catalogs go stale the month you stop feeding them. Connections break quietly. The maintenance is small but it is real, and it is part of the job now, the way keeping a lab notebook is part of research rather than a distraction from it.

The second is that none of this is about the specific tools, and I mean that structurally, not as a disclaimer. My landscape is a set of public commercial assistants and the agentic desktop tools my employer and the market provide. A colleague with different constraints could build the same layers on fully local, open-weight models running on their own hardware. Different vendors, same architecture: an environment that survives beyond one message, an identity, a brain, connections, procedures, a loop with a supervisor. The tools are rented and replaceable. The layers are owned. When I switched primary writing models last year and my voice disappeared from the prose, it was the layers, the accumulated calibration, the catalog, that let me recover in weeks instead of starting over. Whoever owns the layers can change tools. Whoever owns only tools starts from zero every time.

## Where the time came from

So when someone asks where the time came from, the real answer is that I stopped spending it in the one place most people spend all of it: at the bottom of the stack, retyping context into a system that forgets me between sessions. The time was always there. It was being consumed by the absence of structure.

If you are a product manager, an analyst, a domain expert, anyone whose value is judgment applied to a body of knowledge, the instruction is one afternoon long: set up an environment that persists. Then spend the hours teaching it who you are, and the day it takes to catalog what you know. The tools will keep changing under you. The brain you build is yours, it compounds, and nobody can buy it on the same day as their competitors.

*This article is part of an ongoing series on AI product design, healthcare data, and the human side of technology adoption.*