3. The Journey
I did not set out to build an ontology. I set out, like most ambitious people, to feel one step less behind. Nineteen months of MyFitnessPal data, four years of Apple Watch sleep data, a journal I had opened a thousand times. I had more information about myself than any doctor, coach, or therapist I had ever paid, and I could not use any of it. It was data without grammar.
The first thing I learned was that the problem was not more data. The problem was that I had no vocabulary for what I was looking at. I would see the pattern — "I think I do better work after good sleep" — but the moment I tried to act on it, or delegate the instinct to a coach or an AI, the instruction dissolved. "Do better work" is not a thing you can hand off. "High sleep (Apple Watch ≥ 7h) precedes high ambition entries in the next 48 hours 65% of the time" is. The first one is a feeling. The second one is an axiom.
I spent a season learning what the right shape of a personal truth looked like. When I found it, four axioms fell out of my own record almost immediately:
- Learning Master Key (0.67 confidence). High-learning weeks predict higher affect, ambition, and insight the same week. Learning is not a feel-good category. It is the single highest-leverage lever I have on every other domain.
- Exercise × Sleep Synergy (0.57). Neither exercise nor sleep alone drives excellent stress-recovery weeks. The combination does. One without the other is a weak signal. Together they are the signal.
- Belief → Entertainment Lag (0.60). Weeks where I engage heavily with what I believe in predict higher entertainment engagement two weeks later. Depth, then play. Play without depth leaves me flat.
- Zero Negative Impact (0.95). No life domain in my record shows a strong negative correlation with any other. There are no secret costs. Investing in any one of these domains does not quietly take from another. This is the axiom I did not believe until I saw it.
The fourth one changed everything. "Zero negative impact" is not a platitude in my data — it is a 95%-confidence pattern across nineteen months. It means the self-talk that says if you work harder at X you'll lose Y is, in my specific life, wrong. That single axiom reorganized a year of decisions.
The meta-insight was bigger than any individual axiom. The reason I could not previously delegate, hire well, or use AI effectively was not a judgment problem. It was a vocabulary problem. I had the judgment. I did not have the words. An ontology is the first place an ambitious person can keep their own words in a form that software — and other humans — can act on.
That is what Understood.app is. It is the machinery that turns a lifetime of personal data into a working vocabulary, and then uses that vocabulary to answer the next question you ask.
4. The Method
The architecture is eight steps. It is the same architecture whether you are building an ontology, a business, or a body. It works because it is about finding the structural edge and letting it compound, rather than outworking the problem.
1. Context. Accept the data you have. Not the data you wish you had. For me it was nineteen months of nutrition, four years of sleep, a messy journal. For you it is something else. Start with the record that already exists.
2. Circle. Watch before you move. The first two weeks inside Understood.app are diagnostic. You record, the app tags, nothing is predicted yet. You are giving yourself the gift of looking at your life before trying to change it.
3. Close the gap. Find the specific vocabulary you are missing. Most people are missing only two or three words. For me it was "axiom" and "life domain." Once you have the words, you can describe what you want to the software, to your coach, to yourself, and be understood the first time.
4. Choose success. Pick one measurable outcome that, if it moved, would make the rest of your life feel different. One number. Not five. The ontology pressure-tests whether the axioms you are discovering actually predict movement in that number.
5. Code the pattern. As soon as an axiom reaches confidence, put it into the rules the assistant uses. This is the moment the software stops being a journal and starts being an advisor. Code the pattern once; let it run forever.
6. Create a kill switch. Every axiom gets a condition under which you stop trusting it. If your life changes — new job, new city, new body — old axioms expire. The software flags them, you retire them. Truths are not eternal; they are time-bounded.
7. Clear sign of success. Look for the signal, not the feeling. In my record the signal was a Weekly debrief that, three weeks in a row, correctly predicted where my week was headed before I did. That is not vibes. That is a model that works.
8. Compound. Do not overwork the problem. Once the ontology is running, your job is to keep recording and let it improve. The axioms sharpen every week. The recommendations get more specific. The predictions get less wrong. This is the part most self-improvement projects never reach, because most projects are built on willpower, and willpower does not compound. An ontology does.
5. The Invitation
This is not a buy-this message. It is a build-yours message.
Understood.app is the software that turns the record of your life into a working ontology — your own axioms, in your own words, with real confidence scores, used by an AI assistant that finally knows what you actually mean. You get one week of full access for free. If it earns the feeling, you keep it for $100 a month or $999 a year. If it does not, you export everything in one click and leave. The record is yours.
The first hundred people through the door are co-builders, not customers. You get three months free, a direct line to me, and a seat in the room where this gets sharper. The target is not a hundred users. The target is the first hundred people who feel, by week two, that the lights are on.
If you have been one step below where you know you should be, and you are done waiting — this is where that ends.
Read next: The four axioms from my own record · How the ontology works under the hood · Start your 1-week preview