How I think

1. 01

   A probability is a bet, not a fact.

   It lives in your belief, not out in the world. The only honest move is to keep updating it as evidence arrives — the agent at the center is you, and a good system just makes your updates faster, cheaper, and more honest.

2. 02

   The model is the cheap part; judgment is the scarce part.

   Everyone has the tokens now. What stays rare is knowing which question to ask, what a number is allowed to mean, when the model is wrong, and what it costs to be wrong. I sell the calibration, not the compute.

3. 03

   Statistics doesn't lie — the telling does.

   The mathematics is sound; the lie lives in the gap between a rigorous result and how it's told. The caveat you drop is the lie you tell. Be as simple as you can, then stop right before you start lying.

4. 04

   Simple is dense.

   Real simplicity isn't the absence of complexity but complexity compressed without loss. The Central Limit Theorem and the Fundamental Theorem of Calculus are each a sentence long and bottomless. That's the standard.

5. 05

   The average hides the person.

   A summary statistic can hide anything — including a dinosaur. The mean is where you start asking questions, never where you stop. Weight what matters; engineer the features the average erases — the trend, the consistency, the whole trajectory.

6. 06

   Formulation over solution.

   The creative act is asking the right question in the right space; once a problem is well-posed in the right concepts, the answer is often forced. The mere formulation is more than half the work — and finding the space where a problem becomes natural is the master move.

7. 07

   Data is more than what we store.

   If the power goes down, the disk's worth of it is gone — but the real signal often lives in the thought no one wrote down and the timing nobody logged. Measurement is a costly choice, so I stay honest about everything I had to leave dark.

8. 08

   The most valuable model is in your people's heads.

   The plant manager who knows a cell drifts in the cold, the nurse who knows which room runs long. That tribal knowledge is perishable and unauditable. The work is getting it out of heads and into something the team can trust — with their trust intact.

9. 09

   Adoption is the metric.

   A pipeline or analysis succeeds only when people actually use it and it changes a decision for the better. A dashboard nobody opens twice isn't finished. If your team isn't reaching for it, it isn't done.

10. 10

    The perfect company didn't fire its people.

    A system that optimizes the wrong objective will rationally decide its own survival wasn't cost-effective. The good version aims its people at the decisions worth a human and automates the rest — cost-effective in the only timeframe that matters: long enough to still be here.

11. 11

    High rigor is what buys honest simplicity.

    Choice, care, and a little earned complexity are exactly what let you compress without lying. You can't simplify what you never understood. Nuance isn't decoration; it's what keeps the simple version true.

12. 12

    The work is for the porch.

    Less time wrestling data, more time on what actually matters — the people and the open water where judgment lives. That's not a slogan tacked on the end. It's the whole point.

The practice underneath

The move all of this turns on is calibration — and calibration is trained, not declared. “Judgment is the scarce part,” “simple is dense,” “the average hides the person” — each rests on a slower practice off the page: sitting with attention, working honestly with the tools, and learning on purpose. The rotation:

Ānāpānasati — the full awakening of the breath

One of those deserves spelling out. The Ānāpānasati Sutta (MN 118) lays out sixteen contemplations of the breath in four tetrads — body, feelings, mind, and the nature of things. I came to these steps through Thich Nhat Hanh, whose translation and commentary in Breathe! You Are Alive render them as the full awareness of breathing. It is the same move as encoding a subject: one anchor — the breath — and everything else hangs off it, in order.

Four tetrads, sixteen steps: steady the body, know the feeling, work with the mind, then loosen the grip on all of it. It is the calm the fast decisions rest on.

A note, in gratitude. This practice isn't mine — I'm a guest in it. The terms in parentheses are Pāli, the language of the Theravāda canon that preserved the Ānāpānasati Sutta; the English follows Thanissaro Bhikkhu's translation, linked above. I found my way here through living and reading rather than scholarship — Thich Nhat Hanh most of all, and texts like the Heart Sutra along the way — and I keep it here with respect for the tradition that carried it this far.