My Year of Linux on the Desktop: Why Linux?

Preface. Dear reader, I have read this piece through a few times while editing. I am going to sound wholly more paranoid or alarmed than I feel on a daily basis. You may chose to imagine me writing this with a tinfoil hat perched jauntily upon my head.

I want to write about an experiment I’m hoping to run in 2026: shifting a lot of my personal computer use from a laptop to a desktop, and also shifting that machine to a Linux system.

Today, I want to talk about why I'd leave Port Apple, where I've been docked for the last decade. And I want to talk about why Linux on the Desktop ended up as the most serious alternative in 2026.

I’ve been running macOS for essentially the whole time I’ve been writing software. I got a Mac Mini and a MacBook Air halfway through my university career. I’ve used macOS for every “real” job I’ve ever had. I love the aesthetic of typing on a backlit MacBook keyboard in a dark room at night, I love cool little menu bar apps, I love that I’ve always had access to a unix terminal. I even stuck with Apple through the valley of the shadow of death that was the Touch Bar MacBook era.

From a hardware perspective, I don’t know if we’eke ever had it better than we do now. The Apple Silicon chips eek out double-digit hours of battery life from portable devices and I rarely, if ever, find my Apple devices to halter at what I ask them to do (photos, software development).

The story from a software side is a bit different, though. Apple haven’t performed a triple-somersault with the rollout of iOS 26 and macOS Tahoe. Better thinkers and writers than me have documented the troubling design changes.

I hear things aren’t very different on the other side of the aisle. Microsoft’s Windows platform has been rolling out adverts in the menu bar of their (admittedly beta/insider build) desktop operating system. They've mandated what appear to be unnecessarily high hardware requirements for the latest version of Windows 11. Meaning that about 400 million computers currently running Windows 10 will no longer receive free support, though. It's okay, they're letting people pay to receive extended support to run the same operating system on the same hardware that they have been doing this whole time.

And where are they going to go? Where are all those physical computers going to go? On a more personal note: where are the people who rely on those computers for their day-to-day lives going to go? What about organisations like schools or healthcare organisations which run fleets of low-cost computers? Microsoft isn't acting like a company that cares about people or the planet they have to live on for the rest of their lives.

Behind all of this looms the spectre of AI. The promise of generative AI technology to increase productivity (and therefore profits), and a widely held belief that this is a "winner takes all" market has led to a lot of horses standing behind carts asking if maybe we got ahead of ourselves. Essentially all of the biggest tech companies in the world are being asked to show how useful AI is by their shareholders, and by the wider stock market.

Unfortunately, we just don't know how useful, and when, AI is for productivity.

Trillion-dollar companies are doubling down on how central Artificial Intelligence is to their core business model. What was, until very recently, Microsoft Office is now Microsoft 365 Copilot app. Which seems a funny thing to say when the users of your core products (Windows, iOS) are vocally telling you how much worse things feel, or how concerned they are about privacy, or how little reason they see for the change. Not the usual grumblings of “things have changed and I don’t like that” but “sorry, please, you’re getting in the way of my job”.
Hundreds of billions of dollars are seemingly being passed around between companies who build infrastructure to house the computational power for AI services and the companies reselling the output of that compute. Not only are we shockingly uncertain if the AI companies are profitable, but deals are being announced are walked back four months after they should have closed, and also investors seem less confident that "AI investments are generating revenue benefits" (Goldman Sachs), and more concerned that "capital expenditures are outpacing the earnings potential of the new technology" (the FT).

People are spending a lot of money to buy parts of companies which are spending even bigger amount of money to do AI, and we're less confident that those companies are going to be able to turn that bigger piles of money into the biggest pile of money.

The cramming of AI into every surface of a computer might hit harder for Microsoft users (customers?), but something that makes me feel similarly uneasy is the need for the CEO of a company to perform rituals in support of political powers. I am talking about how Tim Cook gave the President of the United States a lovely little gold statue, and then found the company they run exempt from fiscal penalties that all its competitors faced. While I encourage you to consider what that says about the state of geopolitics, I want you to think more specifically about what it means for you. You who will probably never be the giver or receiver of a little golden one of one statue, or someone with the power to directly enact fiscal policy for a nation state.

I did not like, and still don’t, sitting downstream of system where actions like that decide if a company is going to do very well or extremely badly. There is room for nuance here: Cook may have made a masterful performance and read the room flawlessly, ultimately getting what’s good for his shareholders; or he may have bent the knee and revealed how he is nothing but a man devoid of conviction or courage, and that business is all about money. It may even be a little bit of both.

Is Cook always going to be as good (or bad) in his execution of this skill, or as depraved in his ethical shortcomings (or pragmatism)? Is he going to be the CEO at Apple indefinitely?

Are the rules of the game that big companies like Apple are playing (because they are forced to, or because they want to) going to be as predictable, fair, or reliable as they have broadly been for the thirty years before today? The rules that, despite being changing and imperfect, nevertheless let tech companies accrue billions or trillions of dollars in value (and also form monopolies). I don't see much evidence that we're at the tail end of a headily anti-capitalist era, and that finally we can really see what unregulated markets can do.

I thought about how the companies who design and build the software I use every day can crush (through monopolies), profit from (through transaction fees), or regulate (through self-drafted content guidelines and contracts) the marketplaces that would allow me to access some alternatives.

My phone and my computer are basic digital tools that I need to participate in the social and economic world. I don't think I can move to a cabin the woods, to be honest with you. I think I'm stuck here, trying to be a valuable member of my community, and an active participant in the economy. Also I don't want to - I live here, this is my home.

I don't think any of the forces I am describing here will last forever. But big companies run by men (and yes, it is mostly men) with fragile egos and have laid out unfathomable amounts of money and human hours to make AI happen. They have valiantly faced uncertainty and risk, and grappled with the fear that their company might be only the second most rich company in human history.

And I want you to think about who is going to bear the cost of that risk, and who is going to be rewarded when (or if) it pays off?

I sincerely do not believe that I, personally, individually, as a regular old human, am going to be unaffected by these. And I don't think my life is going to be made better.

What's worse - I don't think I'll have a choice in how I'll be affected. It will just sort of have to happen to me. I think things are at best going to get more expensive, and at worse are either going to go away or get worse.

All of this is to say - looking around today, with my current Apple Hardware passing five years of daily use, maybe I don't want to replace it with just another mac mini. Maybe moving my critical daily software use to a platform with lower switching costs and much less corporate interest... maybe that's an idea worth exploring.

Maybe it is time to run Linux on the Desktop.