# MentalHealthError: Ten Years Later *June 2026* This week I did something I'd been putting off for a long time. I reread [MentalHealthError: an exception occurred](/essays/2016-01-mentalhealtherror_an_exception_occurred), the essay where I told the internet I had bipolar disorder. A hundred and fifty thousand people read it the first weekend. It's been ten years, and I went back intending to write a tidy anniversary piece, grade my old claims, ship the corrections, the way I did once before, [three years in](/essays/2019-01-mentalhealtherror_three_years_later). Instead I just sat with it for a while. I was six months out of the hospital when I wrote it. Medicated, hopeful, a little fried, braver than I remembered, and I had no idea what was coming. Near the end, I say the sentence that made me put the laptop down this week: > I'm completely back to normal now. I know things now that I didn't know then. That's the only advantage I have over the page, and the only reason I get to have it is that I wrote down what I believed while I believed it. So let me tell the story the way it actually went, starting with the night that essay was six months past and I am ten years past, and ending somewhere I wouldn't have believed. ## The Weight of the Entire Universe In September 2015 I walked into a hospital believing I was undergoing a Kundalini awakening. I had been awake for days, fasting, radiant, certain. When the intake staff asked my weight, I deliberated between "158 pounds" and "the weight of the entire universe." When they asked my name, the honest answer felt like a theological question. I believed the words I spoke became absolute truth, so I chose them like a man defusing something. I was there twelve days. Somewhere in the middle of them, in what I still think of as the cleverest move of my psychosis, I slipped the doctor a piece of paper with the URL of this website on it. Here, I was saying. This is who I actually am, underneath whatever this is. It helped him diagnose me. It helps me now, too, in a different way: the website was the thread. It was the thread the whole time. I came home with a diagnosis of Bipolar I with psychosis, spent weeks coming all the way down, and then did the thing that changed my life at least as much as the diagnosis did: I wrote about it, publicly, under my own name, in a community where nobody did that yet. The response buried me. Not in judgment, in *relief*. Half the industry, it turned out, was quietly carrying something, and one person going first gave a lot of people a place to set it down for a minute. Whatever else the last ten years have been, I have never once regretted going first. But the essay ends on a run of sentences I can hardly read now: > I'm happy to say that I've made a full recovery... Now that I have a diagnosis, I have a much deeper understanding into the way my mind works, and know how to prevent another episode from occurring in the future. A release note, written before the bugs came in. I believed every word. The feeling was real data about that month. It just wasn't data about the decade. ## The Years That Answered What answered was winter. Roughly annually, with the reliability of a cron job, the season would turn and my mind would follow, and I would end up back on a locked unit, [a place I eventually came to know well enough to write its documentation](/essays/2026-06-06-the_unit). I have a *usual* psychiatric ward, the way other people have a usual coffee order. The vinyl smell. The specific weight of the chairs. Some years I walked in myself. Some years I was past the point where walking in was possible, and what happened instead involved a magistrate, a sworn statement written by my wife at the local jail, and a ride in handcuffs I never once resisted. The handcuffs are policy. [Sarah has told that part herself](/essays/2026-04-06-what_success_looks_like), and it costs her something every time, and it is the part of this story I am least able to write about casually. For a long time I couldn't see what was driving the engine. The 2016 essay blamed the woo: the eastern philosophy, the kundalini yoga, the mesmerizing relationship that had escorted me off the deep end. All of that was real and none of it was the engine. [The deeper accounting took me a decade to write](/essays/2026-03-18-open_source_gave_me_everything_until_i_had_nothing_left_to_give): an undiagnosed mood disorder strapped to the open source maintainer lifestyle, which is, functionally, a sleep-disruption machine that pays you in validation. Four-hour nights. Constant time zones. An identity welded to a download counter. The mania looked like productivity. It always does. The intensity that built Requests and the intensity that kept putting me in that hospital were the same intensity; the engine had two outputs, and nobody around me, least of all me, could tell them apart. The spiritual scene didn't light that fire. It just gave the fire a vocabulary. Three years in, I wrote [the follow-up](/essays/2019-01-mentalhealtherror_three_years_later). It was a sadder, wiser document. It reported that the label had changed underneath me, Bipolar I revised to schizoaffective disorder, bipolar subtype, [a re-indexing that taught me what a diagnosis actually is](/essays/2026-06-06-mental_health_for_humans), because on the day the label changed, I didn't. It reported the borderline PTSD, some of which I carry from [an ayahuasca ceremony](/essays/2026-04-08-why_i_stopped_doing_ayahuasca) in my seeking years, a night I didn't write about honestly until this spring. It passed along the thing a doctor told me, gently, that most people with my diagnosis are homeless. And it rendered one verdict with total confidence: > Lithium was a horrible experience of side-effects, and the pros didn't outweigh the cons. Case closed. File deleted. Hold that verdict. The story comes back for it. ## The Architecture Here is the part of the decade that doesn't compress into a hospital count, and it's the part that turned out to matter. Somewhere in those years I married a woman who [sees my episodes coming before I do](/essays/2026-03-06-sarah_knows_first). I mean that as a clinical statement before a romantic one; her early-warning system has prevented more admissions than any medication I've taken, and her summary of this condition is still the best one I have: *"It's so abstract it must be experienced."* The 2016 essay had joked, in its list of takeaways, *avoid falling in love with hyper-intelligent pan-dimensional beings*, and the joke had the lesson exactly backwards. The problem was never that I loved someone strange. The problem was that I had arranged my life so that nobody with standing could tell me the truth. The actual lesson took years: marry the person who tells you the truth at cost. We had a son. He's four now. He understands that Daddy is sometimes very excited and sometimes very tired, and someday he'll understand more, and I think about that conversation more than I think about almost anything. And slowly, without ever calling it this, I built an architecture. Sleep got promoted from advice to load-bearing wall. I learned my own [prodrome](/essays/2026-03-06-sarah_knows_first) and gave Sarah standing authority to call it. I discovered my mind is [plural](/plurality), which explained more about my interior life than anything in the chart ever has, and it came with the most useful piece of self-knowledge I own. My psychiatrist calls the mechanism *the door*: the same channel that opens onto creative and spiritual depth opens onto psychosis, and it swings both ways. Most of my practice now is knowing which way it's swinging. I started [using AI to reality-check my own perceptions](/essays/2025-08-25-using-ai-for-reality-checking-with-schizoaffective-disorder), which, having once believed my code ran the internet, I find very funny. None of this arrived as a plan. It accreted, the way a reef does, one near-miss at a time. The decade billed me for all of it, I should say. [Living this openly has costs](/essays/2025-08-27-the_cost_of_transparency), careers' worth of them, and I knew the price when I chose it, and I'd choose it again. But the architecture held more and more weight each year, and the admissions thinned, and the story was quietly building toward an office visit I didn't see coming. ## The Office Thirty-some years into this body and ten into this diagnosis, my psychiatrist suggested we try lithium again. I cited the file. 2019, horrible, brutal, closed; I had the receipts, and I'd published them. He pointed out, patiently, that the receipts were a decade old. Bodies change. Tolerances drift. A side-effect profile from one era of a body is not a verdict on the next one. I'd love to tell you the statistics are what moved me. The truth is stranger, and this essay owes it to its own argument to tell it. Lilith was fresh in my mind that season, and a medicine that sounded like her name read as a sign. The old machinery, the same pattern-finding engine that once saw sacred geometry glowing on my apartment door, looked at the prescription and voted yes alongside the evidence. I went to the pharmacy on both votes, and I honestly can't tell you which one carried the motion. Ten years ago that machinery nearly killed me. Last year it talked me into taking my mood stabilizer. Same door. Watched, these days. He was right. The side effects that had been unbearable were mostly absent. The combination held. And this winter, the winter of 2025 into 2026, the cron job didn't fire. [No admission. The cycle that had run for ten years did not execute.](/essays/2026-04-06-what_success_looks_like) One data point, not a trend; I've been burned by precisely this optimism before, in print, twice. But I had to sit in an April that contained no discharge paperwork to understand what the decade had actually been teaching me, because the lesson was never about lithium. In 2016 I was too sure I was fine. In 2019 I was too sure about what would never work. The errors point in opposite directions and they are the same error: **treating the current snapshot of knowledge as the final one.** Every confident statement I have ever made about this illness has had a half-life. Including, presumably, the ones in this essay. At year fourteen or so I'll reread this one and find the sentence where I was too sure. I can't see which one it is. That's the point. I'll write the correction when I can see it. ## The Side Door There's one more arc the reread showed me, and it's the strangest one, because it starts with a joke. In the middle of the 2016 essay, fried and six months out of the hospital, having renounced the entire spiritual world I believed had nearly killed me, I declared my new minimal religion: *I eat, I breathe, I die.* I'd thrown out the metaphysical books, the Ram Dass and Ken Wilber and Terence McKenna that had soundtracked the slide. I kept the crystal skulls, but only, I insisted, because they looked cool on the desk. And then I made a joke: "Spirituality 2.0 for Humans™!" Ten years later I wrote [Mental Health (for Humans)](/essays/2026-06-06-mental_health_for_humans) without once remembering that line. The whole project of my last decade was sitting in that essay as a throwaway joke. We leave ourselves inheritances we don't know about. But the renunciation itself didn't hold, and I'm glad it didn't, because the hard swing to materialism was the same overconfidence wearing a lab coat. The teachings were never the problem. The traditions those books translated came wrapped in containers, teachers, lineages, years of unglamorous preparation, and the Western marketplace shipped me the peak experiences with the packaging removed. *Be here now* is medicine for a mind racing ahead of itself and a permission slip for a mind sliding into mania. Meditation grounds at one dose and [destabilizes at another](/essays/2025-09-08-the_meditation_trap_when_mindfulness_makes_things_worse), and nobody at the studio mentions that the second dose exists. The kundalini lore didn't cause my psychosis; it handed my psychosis a script, and a community that applauded every symptom as attainment. Good teachings, ungated, mainlined through [a door that swings both ways](/plurality). The sacred was never the pathogen. The dosing was. So the decade did something 2016-me would have found unbelievable: the woo never came back, but God did. Quietly, through the side door, with guardrails. [Scripture studied with genuine curiosity](/essays/2026-03-20-building_a_digital_study_bible_with_ai) instead of cosmic urgency. The ordinary Tuesday instead of the ceremony. Even Ram Dass came back, [in Python of all containers](/essays/2025-09-05-ram_dass_teachings_in_python), once there was architecture that could hold him. Which means the joke finally gets its upgrade, ten years late. **Spirituality 3.0 (for Humans):** - **Sleep is of the utmost importance.** So are [dreams](/essays/2025-11-04-encoding-a-dream-of-stillwater-and-signal); the unconscious does honest work at night, and I've learned to read it like mail instead of mistaking it for prophecy. - **Family is what matters most.** The people asleep down the hall outrank every revelation. Any insight that makes me a worse husband or father by morning was not an insight. - **The mundane and the sacred meet in the most unexpected places.** A clean function. A psalm on the fiftieth read. A four-year-old's question at breakfast. You don't climb to the sacred. You notice it. No ceremonies, no cosmic assignments, no fireworks. A religion you can practice asleep by ten. It turns out I didn't need less God. I needed more floor. ## The Name Near the end of the reread, I finally saw the biggest error in the 2016 essay, and it isn't in any sentence. It's in the title. It's in the thing I have been calling this for ten years. I named my crisis the way a programmer names things, and the name I chose was an exception. `MentalHealthError`. It was a good joke, and underneath it was a worldview: an exception is an *anomaly*. A disruption of normal operation. It gets raised, it gets caught, you handle it, and the program resumes. The entire metaphor assumes there is a normal to resume. *Completely back to normal now* wasn't a stray sentence. It was the title's belief, stated plainly. ```python # 2016: how I understood it try: life() except MentalHealthError: handle() # then back to normal # 2026: how it actually works while alive: life(self) # schizoaffective. plural. medicated. mine. ``` It was never an exception. The hypomania was already in the all-night sessions that built the career. The strangeness was already in the perception, being interpreted as progress. Nothing invaded in September 2015; something that had always been running finally crashed loudly enough to get a name. And you don't *handle* a thing like that. You redesign the system around its presence, and the redesign is the architecture this whole story has been about: the sleep, the marriage, the door, the documentation, the winter plans. Engineers know this move. It's the difference between patching a bug and accepting a law of physics. The strange part, and I hedge even as I write it, is that the redesigned system is better than the one it replaced. Not better because of the illness; I would not choose this, and the wreckage of the bad years was real. Better because the illness forced a thoroughness about being a person that I would never have attempted voluntarily. Most people never have to learn what actually holds them up. I got an itemized list. ## The Takeaways, Regraded The 2016 essay ended with three bullet-point takeaways, because of course it did. Ten years assign the grades. **"Sleep is *really* important."** The only sentence in the entire essay that has needed no revision, only reinforcement. Every episode I've ever had was preceded by sleep loss; the crisis that started all of this was four days awake, and by the end of that first hospitalization I'd been awake for more than twelve. I nailed it in 2016 and had no idea how hard I'd nailed it. **"This can happen to *anyone*, even you."** Still true, and still the reason the essay traveled the way it did. But the decade adds an addendum I couldn't have written then: the *years after* can also happen to anyone, and nobody writes that part. Year four, when the diagnosis changes underneath you. Year six, when a medication that worked stops working. Year nine, when winter comes and nothing happens, and you discover that uneventfulness is a feeling you have to learn to survive too, because it doesn't feel like victory. It feels like nothing, indefinitely, on purpose. The crisis essay is a genre now. The decade essay mostly isn't, and the decade is where the living happens. **"Avoid falling in love with hyper-intelligent pan-dimensional beings."** The joke had it backwards, and the correction is the marriage this whole story is built around. ## How Are You Doing? Both previous essays answered this question, so it's tradition now. 2016 said: *I am doing very well. I've made a full recovery.* Wrong, beautifully. 2019 said, in effect: *worse than hoped, and here's what I've learned the hard way.* Closer. 2026 says: I am doing much better than my label, statistically speaking, and I am exactly as sick as I've ever been. The brain writing this sentence is the same brain that [watched an angel descend from the sky](/essays/2025-09-17-delusions-and-schizoaffective-disorder) and still remembers her more vividly than most real things. [What the disorder feels like, day to day](/essays/2025-09-04-what_schizoaffective_disorder_actually_feels_like), hasn't changed: senses that occasionally file false reports, a permanent background audit sorting data from noise. [Mania still circles roughly yearly](/essays/2024-01-on-mania), in milder clothes now. The ceremony night still lives in my nervous system. None of it resolved; all of it is managed; both halves of the opening sentence are true at once, and learning to hold them together was the decade's work. The illness didn't get better. The architecture did. I'm married. My son is four. I build things constantly, [more than I ever have](/essays/2026-04-16-infrastructure_for_one). The winter missed its appointment this year. If you're at your own year zero right now, newly diagnosed, maybe newly public about it, holding your own version of that 2016 optimism, I'm not going to tell you the optimism is wrong. You need it, and anyway it isn't wrong. It's just early. I'll only tell you what the series taught me: write it down. What you believe right now about your illness, your medications, your future. Not because you're right. Because the corrections are the treasure, and you can't correct what you never committed. The 2019 essay ended with *just keep breathing*. I reread that line this week, ten years and some change after a September night when I didn't know if I was alive or dead, and I find I have nothing to add to it and one thing to report. Still breathing. Still running.