Terry Davis, the contrarian [Part I]
Key lessons learned from an outlier who defied all norms

Complete index
- Rabbit Hole
- Attempt at biography
- Utilitarianism
- Standing on the shoulders of giants: side effects (coming soon)
- Complex vs. simple (coming soon)
- Weird and brutally honest (coming soon)
- Creating as medicine (coming soon)
- “Too weird to live, and too rare to die” (coming soon)
- Important resources (coming soon)
Rabbit hole
Last August I went to Patagonia for a week with several friends, also remote software engineers. Since 2022 we have repeated the ritual every year, varying the destination city (Bariloche, Villa la Angostura and, this time, San Martín de los Andes). The plan is usually to work early in the morning and then use the free time for trekking in the mountains and forests, skiing, eating asado, drinking wine and other typical plans of southern Argentina.
One of those days we were killing time in the living room of the house we rented, watching videos on YouTube. It’s not hard to guess what kind of content six developers are interested in. One of them started digging through the videos on Fireship, a programmer’s channel where he shares tutorials, explains fun facts about the industry, gives his opinion on new libraries, interprets technology news, and so on. He selected one called TempleOS in 100 Seconds.
What followed was strange: a boy with an elongated face, jaws present, always wearing a cap, was streaming a game consisting of a group of people in the middle of a desert running in random directions, a rudimentary version of Paint, a program apparently intended to create music, consisting of a keyboard and a score… all in 8-bit user interfaces that reminded me of my first experiences with computers.
The narrator of Fireship gave a rational context to what seemed pure irrationality: an American programmer, diagnosed with schizophrenia, with an IQ out of the normal range, thinks he is God’s chosen one and develops an operating system 100% from scratch that represents the Third Temple prophesied in the Bible, after the destruction of Solomon’s Temple. It includes kernel, graphics library, games and even its own programming language(Holy C) with its own compiler. It seems to be the only example in the history of computing in which an operating system is implemented from start to finish by a single developer.
That was my first encounter with Terry Davis.
Internally, the same thing happened to me as it did about ten years ago, when I came across a documentary about the Voynich Codex, said to be the most enigmatic book ever written: that tangential thing that resists any attempt at categorization, parameterization, framing. An operating system that is not justified by its usefulness (I don’t think there is a single person in the world who uses TempleOS on a daily basis) but to continue a supposed biblical prophecy. A book written in a language that no one knows or can translate. Little nuggets of gold in the desert of repetition.
I knew instantly that I had to dig more about Terry and TempleOS. I saved the link to the video and jotted down a TO DO for the return trip:
[ ] terry davis + templeOS → rabbit hole
This post is the result of having spent some time descending into said hole. An attempt to share what I think is Terry’s legacy on the topics that occupy my head the most.
Biography attempt

There is an abundance of material about Terry on the internet. It’s easy to spend hours and hours between snippets of his livestreams on YouTube, his forum posts, comments from followers and detractors, articles and podcasts that try to analyze him from dozens of different points of view. TempleOS itself, public and free, is the major evidence to try to decrypt its creator. The common denominator of any available material is that it is impossible to stop being surprised by each new piece of information.
With the idea of making this article self-sustaining goes a small, incomplete chronological list about his life and work:
- Born in ‘69, seventh child of eight siblings. Location: Wisconsin.
- Later moved to California and Arizona.
- Graduated with a degree in electrical engineering from Arizona State University.
- Worked for a few years at Ticketmaster as a programmer. It seems to have been his only professional experience.
- In 1996 the party begins: in Las Vegas, still living with his parents, the first manic episodes appear. He is diagnosed as schizophrenic. He receives a government disability pension. According to him, at that stage he receives a “revelation” from God that leads him to abandon his atheism and start building the third temple for the Creator: TempleOS. He spends a decade pursuing this goal.
- Once the work is finished, he goes into outreach. He appears on Reddit and HackerNews with cryptic posts and videos trying to popularize the use of his operating system. For many years he was ignored (his posts on HackerNews have been delisted), accused of spam, victim of troll groups, until for some issue that is not clear to me (enough perseverance? Mythical aura? Some post of his that clicked with the opinion of the online collective?), he began to amass a large following.
- His illness worsens, he begins to have fierce fights with his parents. Some of them are immortalized **on his channel. He ends up on the street in 2017. His last year he spends it as homeless touring cities in the United States. In several videos he is seen with long hair, disheveled, smoking, philosophizing, sometimes dancing, in squares, train stations, city corners. A nerdy and more modern version of Jack Kerouac in On the road: a camera and YouTube replacing the classic notepad of the globetrotting writer.
- He ends his journey in a small town called The Dalles in Oregon in 2018. He films himself at the entrance of a local bookstore, uploads the video to the channel. A few hours later the local media reports the death of a man with an unknown name, hit by a train. What follows are thousands of calls from his fans to said media, trying to confirm if that John Doe was in fact the prolific developer.
Utilitarianism

Trying to promote the use of the first version of TempleOS (called LoseThos), Terry writes in OSNews:
LoseThos is for programming as entertainment. It empowers programmers with kernel privilege because it’s fun. It allows full access to everything because it’s fun. It has no bureaucracy because it’s fun. It’s the way it is by choice because it’s fun. LoseThos is in no way a Windows or Linux wannabe — that would be pointless.
The fact that, as a user, you have access to everything, makes it a ring 0 operating system. From the point of view of usability (any action performed can trigger serious problems if you are not careful) and security (greater possibility of creating malware in unprotected environments), it does not seem to be an excellent decision: protection levels have a reason for being and that is, among other things, to prevent from the design possible unintentional actions by users. Not having them is to delegate that responsibility to them and, in parallel, to delegate a great power to them. Which, according to Terry, is fun.
Another developer, also a user of OSNews, amplifies his sayings in a post titled Recreational programming with LoseThos . He cites several explanations by the author and concludes with the following:
I’m glad that operating systems like LoseThos still exist, in that it shows that computing can still be a hobby; why is everybody so serious these days? If I want to code an OS that uses interpretive dance as the input method, I should be allowed to do so, companies like Apple be damned.
It is very common in the tech world (and probably in the engineering world in general) to associate value with utility: a piece of software produces value only if it is useful to solve a certain user need. The word processor I am using to write this article is valuable because it allows me to translate thoughts into written text that I can then share with the world. Social networks are great products because they connect us. Calendars, because they organize our routine. Streaming platforms, because they allow us to enjoy entertainment shows. Slack, because it facilitates work communication. Google Meet, because it allows us to organize remote video calls. And so on.
That’s why we tend to react with a certain strangeness when presented with examples like TempleOS/LoseThos. It crunches the structure, the mental model in which we assign importance to something: the author himself is telling us, in our face, bluntly, that the main reason he spent years of his life working on this operating system is… to have fun. The argument behind all the decisions about architecture, APIs, backwards compatibility, etc., is, plain and simple: because it’s fun!
An example of the reluctance to this strange way of thinking about software creation is the first comment you read in the post:
2010-09-12 7:58 pm boulabiar
Why people still losing time creating new OSes that looks similar to 80’s ones ? I they have too much time they can help resolving problems of Linux or any common sub-system. Resolving issues/bugs help many millions of users, recreating the wheel every time is a huge time waste…
Wasting time: anything that does not consist of solving concrete needs or collaborating in that collective task is wasting time. The right decision in these cases is to think about what is most useful for the users and put yourself at their service. Why reinvent the wheel with meaningless creations when we can leverage pre-existing inventions?
I remember the first time I came across reflections aligned with Terry’s arguments about his technical decisions when it came to bringing TempleOS to life. It was in a different context: a high school literature class. We had to read several excerpts reflecting on the discipline, interpret and comment on them. One of them stated unequivocally that literature was useless, that it served no purpose.
My first feeling was one of anger. I read a lot and I was annoyed that someone else was claiming that what I did in much of my spare time was of no practical use. I made a mental list of reasons why I considered reading fiction to be useful (stimulating creativity, improving vocabulary, enabling the recreation of fictional worlds from symbols, learning tangentially about culture, history, etc.) and told the teacher about it. She noticed my dissatisfaction. We had an exchange and she made me realize that I had missed the point of what I was reading. The paragraph was a provocation. Doing useless things constituted an act of rebellion in the face of frantic and constant utilitarianism. Sometimes, wasting time gives respite, it is liberating. Adrián Dárgelos tells about his reading habits:
I read to waste time, because wasting time is a subversive action in a world where you have to consume and take advantage of it.
Or:
[This] is a world where wasting time is frowned upon, because what do you do with time, produce money. What do the richest people in the world do? Buy time. So reading is an anti-productive activity, because it doesn’t produce money.
The point here is not to throw everything useful in the trash and pursue religiously tinged side projects. Or that in the next meeting with your Technical Leader you justify your decisions as a developer by saying that you chose X instead of Y because X is more fun. No.
The point is that, on the one hand, we try to relativize more the equation useful == good, and everything else is left out. There are thousands of examples like TempleOS that probably don’t satisfy real user needs on a global level as well as Windows, Linux or macOS do. But is that the only metric to assess the importance of something? My humble opinion is that no, they deserve recognition from other angles, such as creativity, anecdote, myth, thinking outside the box.
On the other hand, do not forget to have fun, to have a good time. I have the naïve idea that in any company or product, having a good time implementing new functionalities is a great differentiator in terms of the pursuit of purpose and mental health in general. Lard moments of utility with moments of enjoyment. If we are constantly producing things just for the sake of them being useful, we are no different from machines in any way.
In short: remembering to exist, rather than to function. As an example, it only takes a few minutes of reading about the life of Terry Davis to understand that he existed. And he had fun.
In the next post, which I’ll publish next week, we’ll talk about the side effects of aggressively increasing levels of abstraction.