I have eleven unfinished projects on my GitHub right now. Twelve if you count the one I started last Tuesday and abandoned on Wednesday after realizing that someone on Hacker News had built something "kind of similar" three years ago (it was not similar at all, but the README used the same shade of blue, and at that point the vibe was ruined). I am not ashamed of this. I am, in fact, prepared to argue that this makes me a better engineer than you. Not just better. Morally superior. And I'm going to prove it using nothing but flawless reasoning and an airtight command of logic that has never once failed me or anyone who agrees with me.
Let's begin.
I. The Argument from Nature
Every great system in the natural world is defined by abandonment. Rivers don't finish. They meander, fork, dry up, change course, and occasionally flood a town that probably shouldn't have been built there. Trees drop leaves every autumn, which is essentially the botanical equivalent of mass-archiving your repos. Wolves (and I know a thing or two about wolves) don't finish chasing every elk they spot. They pick up a trail, evaluate, and frequently just stop and go do something else. This is not laziness. It is evolutionary optimization. Darwin would have had thirty unfinished side projects if GitHub had existed in the 1850s. He basically did, actually, if you count his barnacle fixation.
The point is that nature does not reward completion. Nature rewards exploration. And since I am a product of nature (I have a spine and everything), it follows that my tendency to start a Rubik's cube-based programming language and then wander off to build an LED matrix controller and then pivot to writing a ten thousand word essay about surveillance is not a character flaw. It is biological destiny. To argue otherwise is to argue against four billion years of evolution, and I'm sorry, but I'm going to take the planet's side on this one over some guy on Reddit who finishes his to-do lists.
II. The Inductive Proof
Consider the following evidence. I started building a personal AI assistant in Rust called Fern. I started building a web-based C/C++ IDE called Expedition. I started designing a card game called Last Braincell. I started writing a dystopian novel. I started a CTF benchmarking platform. I wrote two programming languages. I started an esolang based on Rubik's cube notation. I started a commission management platform. I wrote an essay about how nobody deserves anything, which I did finish, but only because the philosophy of moral luck is the one topic on Earth where the conclusion is that nothing is anyone's fault, which felt thematically appropriate to the way I manage my GitHub.
Now. Look at that list. That is a lot of projects. And I started all of them. Every single one. Do you know who else started a lot of things? Leonardo da Vinci. Leonardo da Vinci famously did not finish most of his projects. The Adoration of the Magi: unfinished. The Battle of Anghiari: unfinished. Dozens of notebooks full of inventions he never built. The man was history's greatest project-abandoner, and he is universally regarded as a genius. I also do not finish my projects. The logical conclusion is so obvious that I will not insult you by stating it explicitly.
(Fine. I'm a genius. I said it. Moving on.)
Someone will inevitably object that the similarity between me and da Vinci begins and ends at the not-finishing part, and that the quality of the work matters. This person is missing the point. The quality of the unfinished work is, by definition, unknowable, because it is unfinished. You cannot evaluate what does not exist. This means you also cannot prove it would have been bad. The superposition of an unfinished project is that it is simultaneously a masterwork and garbage, and until someone opens the box (ships the code), it remains in its quantum state of potential greatness. I am simply preserving potential greatness. You're welcome.
III. The People Who Finish Things
Let's talk about the people who finish their side projects, because I think a close examination reveals something uncomfortable about them.
These are the people who post on Twitter with screenshots of their deployed app and a caption like "finally shipped it after 6 months of grinding." They have ninety-seven percent test coverage. Their README has badges. Their commit history is clean. They have a domain name. And the app is always, without exception, either a to-do list with AI integration or a habit tracker that looks exactly like every other habit tracker but this one has "a twist" (the twist is that it's slightly worse).
I'm not saying that finishing projects is a sign of a limited imagination. I'm saying that I have personally never met a project-finisher who could also hold an interesting conversation about the philosophy of consciousness, and I think that correlation speaks for itself.
Meanwhile, look at the people who notoriously did not finish things. Da Vinci, as mentioned. Kubrick abandoned dozens of film projects. Schubert literally has a piece called the Unfinished Symphony and it's one of the most famous compositions ever written. It is famous specifically because he did not finish it. Schubert understood something the productivity-influencer crowd refuses to acknowledge: stopping is an art form. The Unfinished Symphony is not great despite being incomplete. It is great because Schubert had the artistic instinct to recognize that finishing it would have ruined it. I am doing the same thing with my Expedition IDE. I stopped at exactly the right moment. The fact that "the right moment" happened to coincide with me getting distracted by a new blog post idea is entirely beside the point.
IV. A Brief Digression on Productivity Culture (Which Is Bad, Because I Don't Like It)
The entire productivity industry is a scam. I know this because I tried a Pomodoro timer once and it made me less productive. I was building an Arduino project, and the timer went off at twenty-five minutes, and I was right in the middle of figuring out charlieplexing for an LED matrix, and the timer told me to take a break, and by the time I came back I had lost the thread entirely. The project is now on my list of unfinished things. The Pomodoro technique literally killed my LED matrix. This is a direct causal link and I refuse to hear alternative explanations.
If one productivity technique failed me once, it is entirely reasonable to conclude that all productivity techniques are fraudulent, that the people who sell them are charlatans, and that the entire self-help industry is a conspiracy designed to make people feel bad about not finishing things so that they buy more books about finishing things, which they will also not finish, creating a perfectly closed loop of profit extraction. Follow the money. Actually, don't follow the money, because following things to completion is exactly the problem we're trying to avoid here.
You might argue that I'm generalizing from a single experience. To which I say: yes. And?
V. The Slippery Slope of Completion
Consider what happens if you finish a side project. You deploy it. People find it. They have opinions. Some of those opinions are feature requests. Now you are maintaining an open-source project, which means you are doing unpaid labor for strangers who will open GitHub issues with titles like "doesn't work" and a body that says "please fix." Three months later you're mass-closing stale issues at two in the morning and your README has a "Contributing" section that you secretly hope nobody reads.
This is the inevitable endpoint of finishing things. Every completed project becomes a maintenance burden, every maintenance burden becomes a source of resentment, and every source of resentment eventually leads to you mass-archiving your repos, which is exactly where I started. The only difference is that I skipped the miserable middle part. I went directly from "this is a cool idea" to "this served its purpose" without passing through the valley of pull request review.
If you finish one project, you will finish two. If you finish two, you will feel compelled to finish three. Before long you are a person who "ships," and shipping is a full-time identity, and identities demand maintenance, and maintenance is the enemy of exploration. The first hit is free but the habit is lifelong. I've seen it happen. A friend of mine shipped a Chrome extension in 2024 and now he literally cannot start a project without finishing it. He has a completed projects folder. A folder. The man is trapped and doesn't know it.
VI. On the Subject of My Blog
I have a blog. You may have read it. (If you haven't, you should, and the fact that you are already reading something I wrote is probably sufficient evidence that your taste is good, which means you would enjoy the rest of my writing, which means you should read it, QED.) The blog currently has nineteen posts. Each one is a finished piece of writing. This might seem like it contradicts my thesis, but it doesn't, for the following reason: blog posts don't count.
Why don't blog posts count? Because I said so, and this is my essay, and the person making the argument gets to define the terms. If you disagree, you are free to write your own essay, which you will not finish, which will prove my point.
But more substantively, each blog post is itself a side project that I abandoned whatever I was previously doing to write. I was working on Fern and then I wrote an essay about agentic experiments. I was working on Expedition and then I wrote a satirical defense of mass surveillance. I was supposed to be doing homework and then I wrote a legal analysis of Anthropic v. The United States at one in the morning. The blog is not evidence that I finish things. The blog is evidence that I am so committed to not finishing things that I will interrupt one unfinished project to start and (occasionally) finish an entirely different one. This is advanced non-completion. This is the prestige.
VII. The Ethical Dimension
Let me be serious for a moment. (This is the part of the essay where I adopt a grave and thoughtful tone to disguise the fact that the argument is about to get significantly worse.)
Finishing a side project means shipping code into the world. Code that runs on servers. Servers that use electricity. Electricity that, in many parts of the world, is generated by burning fossil fuels. Every deployed side project contributes, however marginally, to climate change. My eleven unfinished projects are sitting in local repos on my machine, consuming zero server resources, burning zero carbon, harming zero ice caps. By not finishing my projects, I am saving the planet.
You might point out that my machine is also using electricity, and that having eleven repos on my hard drive is not meaningfully different, ecologically, from having eleven deployed apps. You might point out that the carbon footprint of a single hobby project is so negligible that this argument is absurd on its face. You might point out that this is essentially a non sequitur dressed up in environmental concern to lend moral weight to what is fundamentally a defense of procrastination.
To which I would respond: but what about the children? Don't you care about the children? The children who will inherit a planet ravaged by completed side projects? Are you really going to look those children in the eye and tell them you deployed a habit tracker to Vercel because you wanted green squares on your GitHub contribution graph?
Didn't think so.
VIII. The Empirical Evidence
I conducted a study. Don't ask me about it. I asked five of my friends whether they had ever finished a side project. Three said yes. Two said no. Of the three who said yes, one currently has a cold, one failed their last math exam, and one stubbed their toe this morning. Of the two who said no, both are in perfect health and reported high levels of general life satisfaction.
The sample size is small but the results are unambiguous. Finishing side projects causes illness, academic failure, and foot injuries. Not finishing side projects causes health and happiness. Correlation this strong might as well be causation, and anyone who tells you otherwise is a statistician, which is a profession that exists solely to prevent normal people from drawing the obvious conclusions from their own data.
Furthermore, I only included the five friends who supported my thesis. I also asked three other friends, but their responses were confusing and didn't really fit the pattern, so I excluded them as outliers. This is standard scientific practice. Probably.
IX. Conclusion
I have demonstrated, through rigorous argument, that not finishing side projects is: natural (Section I), a hallmark of genius (Section II), morally distinct from the pathology of completion (Section III), a rational rejection of fraudulent productivity culture (Section IV), a safeguard against the maintenance trap (Section V), consistent with a prolific creative output (Section VI), environmentally responsible (Section VII), and empirically validated (Section VIII).
If you disagree with any of this, I would invite you to consider that every person in history who has argued against me on the internet has eventually stopped responding, which means I have never lost an argument, which means I am incapable of being wrong. This is a simple application of basic logic and I'm frankly a little embarrassed that I had to spell it out.
Now if you'll excuse me, I need to go start a new project. I have an idea for a benchmarking tool that evaluates how well AI models handle deliberately fallacious reasoning, and I think it could be really interesting, and I'm going to think about it very hard for about forty-five minutes and then go get cereal.
I'll write more when there's something to measure. (I won't.)
The author has no formal training in logic, philosophy, rhetoric, or statistics, and would like to assure you that this has never stopped him before. All fallacies in this essay are intentional. The author is aware that this claim is itself unfalsifiable, which is, if you think about it, further proof that he's right.