A standing note on process, sources, and where AI fits
Every article on this site is researched, turned over, verified, and edited by a person. That person is me. This page explains how my process works, including the part where I use AI, because anyone who reads what I publish deserves to know how it was made.
What follows applies to any article carrying my byline. This site has been running for years and holds a great deal of older work, which I am steadily rewriting and republishing under my own name and this standard. That takes time and I am giving it the time. If a piece does not show my byline yet, it has not been through this process, and it will be.
Where the Ideas Come From
The thinking is mine, and it comes from somewhere specific. Thirteen years of higher education at the University of Michigan, through college, medical school, and residency. Ten years as a board-certified anesthesiologist, in private practice, in employed practice, and as a locums physician across multiple states.
A family that started with nothing. After college, my father was one of about one percent of candidates in his field to pass a government examination in Taiwan, then saved his own tuition for three years and three months. He left for the United States with two suitcases, for graduate school. My mother was a middle child in a family of ten and was nearly given up for adoption, until her paternal grandmother had the final say. She went on to raise four kids while earning a master’s degree and working full time. I am the first physician in my family, which meant that most of what I know about doing this work well, I had to find out for myself.
Siblings I am still close to. A husband I garden, travel, and go birding with. Fierce devotion to family, the kind you are born into and the kind you choose. A decade that held too much loss and pain, and the understanding that more is coming. A mind that will not stop reading.
And a set of things I will not negotiate: my integrity, the golden rule practiced rather than admired, gratitude for every bit of what I have, and a commitment to people, to the planet, and to what’s possible. I take the things that matter completely seriously and everything else not seriously at all, because life is too short otherwise.
When I write about gratitude, or trust, or what integrity costs, I am writing from all of that. It is the part no tool supplies, and it is what makes an article worth reading in the first place.
I hold onto that carefully. When an idea in one of these articles turns out to belong to someone who thought it first, I say so and I cite them. When it turns out to be mine, I say that too.
What AI Does Here
I use AI as a research and drafting partner. Concretely, that means four things:
- It finds sources. It searches faster and wider than I can, which means I can chase a claim back to the study it came from instead of settling for the article that summarized the article that summarized the study.
- It finds who got there first. Often an idea I believe is mine turns out to have a long history. Knowing that lets me credit the people who came before and then say clearly whether I agree with them or am reaching for something different.
- It sharpens the thinking. This is the largest part by a wide margin. I bring a half-formed idea, and we take it apart until I know exactly what I mean. If my reasoning has a gap in it, I want to find that before a reader does, so I ask for the hard read and I get one.
- It edits. Structure, clarity, grammar, the sentence that has been bothering me for two days. I am meticulous to a fault, and I will rewrite a single line a dozen times before I let it go.
Where the line sits: every idea, every judgment call, every decision about what stays and what gets cut belongs to me. I write under my own name, and I can talk at length about any article here. Why a paragraph was removed. What I was reaching for and did not quite catch. What I changed my mind about halfway through. That is the real test of authorship, and it is one I welcome.
How a Claim Earns Its Place
Accuracy is the whole product. A confident wrong answer is worse than a slow right one, so the standard here is deliberately strict.
Every factual claim is checked against a source I actually retrieved, not something I remembered. Memory is a wonderful thing, but a poor citation.
Primary research comes first, then reputable and government sources. A site that asserts a fact without saying where it came from does not count as a source.
Numbers, dates, superlatives, attributions, and any sentence explaining why something is the way it is get flagged and verified individually. Those are where errors hide.
Press releases are labeled as press releases. They are not the study, and when only an abstract is reachable, I say so.
When the evidence is unsettled, I show you where the uncertainty sits. If the studies disagree, we look at how they disagree and what that means. If nobody knows yet, I say nobody knows. What I will not do is round a shaky finding up into a confident sentence.
Every article I publish is accurate as of the day it publishes. Links are meant to be clicked. Follow them.
I name my sources in the text, so you can find the references even if their URLs stop working. Sites reorganize, links break, and researchers revise their thinking, which is how it is supposed to go. When I catch a meaningful change to something I have cited, I come back and say so. What I will not pretend is that every link here will resolve forever. If you hit one that does not, let me know and I will fix what I can, when I can.
When someone shares knowledge that sharpens my thinking, I am grateful for it, and I will tell you what changed. Thank you for that.
Why My Writing May Carry an AI Watermark
In documentation published in August 2026, Anthropic described how it marks AI-generated content: an imperceptible machine-readable watermark embedded in text its Claude models generate, and signed provenance metadata attached to generated files under the C2PA standard. The mark travels with the text through copy and paste. Other AI companies are doing the same, because Article 50 of the EU AI Act, Regulation (EU) 2024/1689, has applied since August 2, 2026 and requires providers of generative AI systems to mark their outputs in a machine-readable format so they can be detected as artificially generated.
Worth understanding: that mark indicates content was processed by the model, not written by it. Hand it a paragraph you wrote yourself and ask for a proofread, and the paragraph comes back marked. So, a lot of what I publish will carry a signal, including sentences that began in my notebook.
Which is fine, because I am telling you up front. A watermark only threatens a writer who is hiding something.
The related tools, the ones that scan finished text and guess whether a machine wrote it, are a different story. They work by measuring how predictable your word choices are, so careful and well-structured writing scores as suspicious. In a 2023 paper published in the Cell Press journal Patterns, Stanford researchers Weixin Liang, James Zou, and their colleagues tested seven widely used detectors and found an average false-positive rate of 61.3 percent on essays written by non-native English speakers, while the same tools classified U.S. student essays accurately. They recommended against using these detectors to evaluate anyone.
I was born in Michigan; English is my first language. My writing has still been flagged as fully AI-generated more than once, in pieces where no AI was involved at all. The instrument reads zero when it should read one, and one when it should read zero. Disclosure is a better answer than detection, which is why this page exists.
When I Get Something Wrong
I will. When it happens, I fix it, and how visibly depends on what broke.
A typo, a dead link, a clumsy sentence, a formatting slip: corrected quietly. The goal is an accurate article, and nobody needs a footnote about a misplaced comma.
A wrong number, a misattributed idea, anything that changes what a claim says: corrected and noted. Someone may have read it, believed it, or built on it, and they deserve to know the ground moved.
If you find an error, please tell me. Corrections get made whether or not I manage a reply, so send it along regardless. I would rather be corrected than be comfortable.
The Story Behind Any of It
If you are curious about how a particular piece came together, where a statistic came from, or why I landed where I did on something contested, you are welcome to ask. Every article here has a story behind it and I like telling it. In fairness, I should say that I am often slow to answer and never quick on the spot. I have a great deal on my plate, and my brain files things in an order known only to itself.
One Last Thing, About Em Dashes
I like them. I have always liked them, back when nobody found that suspicious. Lately I am told they read as evidence of a machine, so for a while I used them less, which is its own small surrender — writing worse to look “more human.” Since my work will carry a mark either way, I am taking them back.
Why Any of This Matters
a little dose of happy® is about how happiness actually works. Not as a mood that arrives on its own, but as something you can act your way into. It is built on behavioral activation, a body of clinical research showing that small deliberate actions change how you feel, and that you can start before you feel ready.
I write from the life I know: what I have lived through, the vast experience behind me, the reading and thinking I never stop doing, and a deep love for people, for the planet, and for what is possible.
The care I take with sources comes from the same place. I am asking you to try things, and that is not a small ask. I would rather take the time to be right and say plainly when nobody knows yet.
If you found your way here, you probably came with a question. Most people arrive with some version of the same one: how do you stay happy when so much of life is hard? I never expected anyone to ask me that. People do, and it turns out I have a great many thoughts about it, which is why this blog exists.
What I hope you take from it is that it is still okay to laugh, still okay to enjoy the strange wonder of being alive, and that the human parts — kindness, care, empathy, respect, gratitude, integrity — matter more when the world is dark, not less.
I am glad you are here. Stay as long as you like. 🩵
Manda
Co-Founder and Chief Visionary Officer, a little dose of happy®