Photo by J.M Read on Unsplash

Author’s Note

I’ll start by being blunt: This article isn’t a broad-spectrum Microsoft Bad™ rant.

It’s not a rejection of the company, the people, or the vast ecosystem of tools that continue to do meaningful, empowering work. I still actively support and advocate for the Power Platform, Azure, Microsoft 365, and the internal product teams whose focus remains on user enablement, accessibility, and organizational capability.

That work matters, and in many cases, it continues to get better.

But Windows, the platform that once anchored that ecosystem, is changing in ways that deserve scrutiny. This wasn’t an article I set out to write. It came after long observation, deep research, and increasing unease. Because what I’ve seen isn’t just a shift in features. It’s a shift in values.

Where Windows once positioned itself as infrastructure, quiet, capable, and responsive to its users, it now behaves more like a product optimized to extract value from those users. That’s not a conspiracy, so much as a strategy. And it’s one that appears to be accelerating, even as market share quietly erodes.

This piece is also not intended as an indictment. It’s an attempt to name the tension between the platform Windows was, and the incentives it seems to be aligning toward now. My hope is that it invites more reflection than outrage.

For those of us who’ve built careers around this ecosystem… indifference would feel like a betrayal, and honest support currently requires criticism.

TL;DR – For those looking for the “why” behind this transition, this article explores three core shifts in the Windows architecture that have redefined the user relationship:

• The Cost of Telemetry: Why the perceived “heaviness” of the modern OS is not a technical accident, but a functional requirement of a business model built on behavioral data loops.
• From Infrastructure to Product: How the OS has moved from being a neutral stage for work to an active participant competing for user attention.
• From Operator to Subject: The transition from a system that provides tools for the user to one that enforces a specific posture (security, identity, and engagement) through mandatory defaults.

~Dom

My first computer didn’t have windows, just a blinking green cursor, a keyboard, and the ‘soft’ whir of a hard drive spinning to life under MS-DOS. Commands mattered. Cause and effect were immediate. You typed something, and the system responded, honestly, and without agenda.

A few years later, I sat down in front of Windows 3.1 at a relative’s house. Icons, folders, scrollbars… an interface layered on top of interface. It was the first time an operating system felt like something you could inhabit. And from that point on, Windows became a kind of companion. Familiar. Capable. Sometimes frustrating, but rarely indifferent.

I’ve used every major version since then. Windows 95 felt like an event. 98 was refinement. ME and Vista were lessons. XP held its ground long enough to become foundational. Windows 7 earned its loyalty. 8 got lost for a while. 10 brought things back into alignment. And for a time, it felt like the platform was maturing alongside its users; iterative, responsive, and increasingly shaped by participation.

I joined the Windows Insider Program. I filed feedback, tested preview builds, and treated the platform like something we were building together. Updates weren’t just tolerable, they were hopeful. The system had its quirks, but it was fundamentally trying to get better.

Then came Windows 11.

It wasn’t broken. It wasn’t unusable. But something subtle had shifted. The system still ran. The features still worked. But the relationship was different. What had once felt like infrastructure now felt like a product, one increasingly optimized for itself.

That was the turning point. Not a bug, not a crash, not a single update gone wrong. Just a gradual, quiet realization: the OS no longer shared my priorities.

So I stepped back.

I’m not leaving Windows in protest. I’m leaving because the incentives have changed, and the assumptions changed with them. And once you see that clearly, the question isn’t “why leave?” so much as it becomes “why stay?”.

The Platform That Competes for Your Attention

There was a time when paying for software bought you silence. You weren’t just acquiring capability, you were exempting yourself from the noise. Windows 11 has quietly ended that contract.

This is no free-tier platform subsidized by ads. Every consumer machine ships with the cost of Windows embedded in the price, baked into the sticker, licensed as necessary infrastructure. And yet the experience increasingly resembles something designed not to serve you, but to study and shape you.

Promotions are now part of the interface. The Start Menu recommends. The File Explorer promotes. The Lock Screen nudges. Even the Settings app, a space once reserved for configuration, now surfaces offers and engagement cues, framed as suggestions but functionally indistinguishable from ads. Some push Microsoft services. Some point outward. All compete for your attention.

And critically, none of this is opt-in.

These systems are enabled by default. Turning them off, when possible at all, involves navigating a patchwork of toggles, policy settings, and registry edits, some of which are reset after updates or ignored entirely on non-enterprise editions. The controls exist, but they are designed to erode over time. Saying “no” becomes an act of maintenance.

This isn’t how good infrastructure behaves.

Operating systems have historically been neutral ground, tools that make work possible without interfering with it. They surface capability, not suggestion. But Windows 11 has adopted a new posture: one where the OS is no longer just the stage, but a participant with its own incentives.

That’s a meaningful shift. Once the operating system begins optimizing for engagement, it stops being passive. It becomes interested in outcomes. It starts treating attention as a resource to be guided, harvested, and monetized.

That’s what makes this different from simple annoyance. It’s not just about an ad in a menu. It’s about a design philosophy where the user is no longer the operator of the system, or even the customer that already bought the software, but part of its telemetry loop.

In contrast, Linux distributions treat user attention as off-limits by default. Nothing appears unless installed. Nothing persists unless configured. Systems behave like systems: explicit, observable, and content to remain quiet unless asked to speak.

Windows 11 has made a different choice. And while that choice may serve the platform’s goals, it no longer aligns with mine.

Loss of Functionality Masquerading as “Simplification”

One of the quieter regressions in Windows 11 is the disappearance of capabilities that once defined the system’s flexibility. This is often framed as simplification, but in practice, it’s something else entirely: constraint, framed as clarity.

The most visible example is the taskbar.

For decades, it was a surface tuned by the user. Ungrouping windows, shifting alignment, customizing behavior… these weren’t hacks or tweaks, they were features. In Windows 11, those affordances are gone. The new taskbar centers itself, visually polished, but functionally diminished. Preference yields to prescription, and the result is a cleaner interface that does less.

Context menus follow the same pattern. What was once a single right-click is now a two-step process. Familiar options are buried behind Show more, a phrase that translates roughly to the real controls are over here. The underlying capability hasn’t vanished entirely, sure, but it’s been placed out of reach by default, as though users are being gently dissuaded from remembering what used to be immediate.

Settings tell a more fragmented story. The new Settings app feels modern, but incomplete. Core system behaviors remain tucked away in legacy Control Panel dialogs, interfaces the platform can’t fully remove without breaking compatibility, but won’t fully embrace again either. The result is an OS in split-screen: one foot in the past, one hand on a paintbrush, neither side able to finish the job.

To the user, none of this feels accidental. This isn’t simplification in the sense of streamlining complexity. It’s simplification as redefinition: narrowing the range of acceptable user behavior to align with an intended design path.

That’s the shift, subtle perhaps, but deeply consequential. Aesthetic consistency is now privileged over capability or user preference. Predictable workflows give way to curated ones. The designer’s vision takes precedence over the operator’s intent. Deviations aren’t supported; they’re deprecated or patched.

While this is often positioned as a technical choice, we can’t deny that it’s also a philosophical one.

Operating systems were built to adapt. They serve at the discretion of the person sitting in front of them. When a platform begins to enforce how it must be used, or when customization becomes an obstacle instead of an expectation, it stops behaving like a system and starts functioning like a service.

And services, unlike tools, are not always yours to shape.

Performance That Answers to the Wrong Master

Set aside the interface changes, the ads, and the nudges. Even at a purely technical level, Windows 11 introduces friction. It is slower, heavier, less predictable… And not by accident.

On equivalent hardware, the difference is visible. More services run in the background. More abstraction layers sit between the user and the system. Telemetry processes hum quietly under the surface. What was once deterministic now feels reactive, like the system is busy with itself.

These aren’t show-stopping flaws. They’re paper cuts. Small, cumulative, and hard to isolate… at least until you switch machines or operating systems and notice what’s missing: the pauses, the delays, the low hum of unseen processes steering the machine on your behalf.

Windows 10, on the same hardware, feels faster. Not because it’s older or simpler, but because it defers less and assumes less. It runs fewer background negotiations. Linux goes even further. Across the board, from boot time, to idle resource usage, and even I/O consistency, it is measurably faster. Not only in theory, but in real metrics that affect how long you wait, how often you’re interrupted, and how predictably the machine responds.

This isn’t about the kernel, It’s about the system’s priorities.

Linux performs better because it is lighter, not by coincidence, but because it lacks incentives to be heavy. It doesn’t track behavior. It doesn’t surface suggestions. It isn’t built to shape engagement. Its processes exist to serve the user that installed them, not the platform’s business model.

That, again, is a philosophical difference. When an operating system sees the machine as a tool owned by the user, efficiency is a natural side effect. When it sees the machine as a terminal for influence, overhead becomes an investment; an acceptable cost for telemetry, engagement signals, and behavioral scaffolding.

On my LG Gram, Windows 11 booted in about 25 seconds. Arch Linux, on the same hardware, cuts that in half. No hardware upgrades. No tweaks or optimizations just for speed. Just a natural result of the absence of noise.

Once you see that difference, it’s hard to unsee. Performance stops being a technical detail and starts revealing what the system values, and who it’s really working for.

Footnote: On Security and Choice

It would be intellectually dishonest to claim that every background process in Windows 11 is a telemetry-hungry parasite process. A significant portion of the system’s “heaviness” is actually the result of genuine security innovation, specifically Virtualization-Based Security (VBS) and Hypervisor-Enforced Code Integrity (HVCI).

From a systems architecture perspective, these are impressive. By using the hardware’s virtualization capabilities to create an isolated “secure zone” for the kernel, Microsoft has effectively moved the goalposts for rootkits and sophisticated malware. For a fleet of 50,000 corporate laptops, this is a massive win for the defensive line.

However, this security comes with a “performance tax” that is often measured in double-digit percentages, depending on the workload. In previous eras of computing, a trade-off this significant would be a choice left to the operator. You would weigh the risk of your environment against the need for raw I/O throughput or CPU cycles.

Windows 11 removes that choice. By making these features mandatory and non-trivial to disable (often requiring BIOS-level changes or deep registry edits that the OS might “correct” later), the platform has transitioned from providing security tools to enforcing a security posture.

I do not begrudge Microsoft for building a more secure kernel. I begrudge the assumption that I, the person who owns the silicon and bought the license, am no longer qualified to decide how its power is distributed. When “secure by default” becomes “secure by mandate,” the user is no longer an administrator of their system; they are a guest in a managed environment.

Security as a Shibboleth

The cleanest break Windows 11 makes from its own lineage isn’t in the interface, it’s in the hardware requirements. Compatibility, once a core strength of the platform, is now conditional. Machines that ran Windows 10 smoothly for years, stable, secure, and performant throughout, were suddenly declared ineligible for the upgrade. This wasn’t because they failed to function, some of them even installed and ran 11 with workarounds, but instead because they failed a checklist.

TPM 2.0, Secure Boot, and approved CPU generations. These were framed not as preferences, but as minimum security standards. The implication was clear: devices that didn’t qualify weren’t just unsupported, they were unsafe.

In practice, however, the story was different. These requirements were easily bypassed by users who found workarounds. Within days of release, unsupported systems were running Windows 11 without issue. They weren’t running into crashes, instability, or a sudden wave of compromised endpoints. Worse, the new requirements showed no measurable, user-visible security gain that would justify the exclusion.

What followed wasn’t clarity, so much as it was churn. Perfectly viable machines were pushed aside; functional hardware rendered obsolete. The message was implicit, but unmistakable: your device isn’t welcome. It wasn’t because it can’t run the OS, but because it didn’t fit the model.

This hints that the requirements model isn’t about safety, but rather about control. Control over hardware lifecycles. Control over user upgrade cadence. Control over what hardware qualifies as a participant in the Microsoft ecosystem. It’s probably not intentionally malicious, but the outcome is the same: fewer paths forward for older machines, and tighter coupling between software, hardware, and identity.

And when viewed alongside Windows 11’s growing appetite for telemetry and attention, the picture sharpens. Security becomes the justification, but the pattern reflects strategy. It also hints that the requirements weren’t designed primarily to protect users from risk. They were designed to enforce conformity; standardizing the machines that run the platform, not necessarily to improve their safety, but to make them easier to shape.

From a user’s perspective, it doesn’t feel like protection. It feels like disqualification.

Privacy as Default Denial

Windows 11 doesn’t treat privacy as a right. It treats it as a setting, and one that comes preconfigured against you. What was once optional is now assumed. A Microsoft account is no longer a feature to enhance integration, it’s a requirement. What was once local is now cloud-first by design. Integration is the excuse, but centralization is the outcome.

Your user identity is no longer a local abstract. It is fused to the machine, anchored by hardware IDs, reinforced through TPM-bound credentials, and monitored through telemetry that actively resists permanent opt-out. Especially on consumer editions, where privacy toggles don’t stay toggled; updates quietly reset them. Behavior shifts quietly with each release, and your consent becomes a moving target.

The deeper concern isn’t the data collection, so much as the implied collapse of boundaries.

Identity, device, and usage are no longer separate concepts. They’re fused into a persistent profile, one that spans updates, networks, services, and environments. It doesn’t just identify the user on the system. It defines them, across connected machines, networks, time, and touchpoints. That profile, by design, is accessible to the Microsoft ecosystem, and by extension, its commercial orbit.

This isn’t architecture for safety. It’s infrastructure for governance. Systems that want to protect user privacy create and enforce separation between these types of data. Systems that want to leverage it collapse that distance until the user, the machines they use, and the data they generate are indistinguishable.

That difference, too, matters. Privacy isn’t lost through breach. It’s eroded through the design of the system. That’s the only way it can be when the defaults assume visibility, and the opt-outs are buried deep enough that only power users find them.

When the system binds who you are, what you use, and how you behave into a single, portable identity, privacy isn’t a choice anymore. It’s something you’re granted, conditionally, and on someone else’s terms.

When the Pieces Lock Into Place

Taken alone, none of Windows 11’s decisions are indefensible. The TPM requirements can be framed as security. Requiring Microsoft accounts can be sold as convenience. Telemetry can be positioned as feedback. Recommendations as helpful design for novice users. Background processes as the cost of modern software capabilities.

But Windows 11 is not defined by these decisions individually. Like any complex system, it’s defined by how precisely they fit together. A TPM stores your credentials. Those credentials require a Microsoft account. That account links to hardware. That hardware generates usage telemetry; app habits, location data, and engagement patterns. That data feeds into personalization pipelines, which then surface targeted nudges inside the OS itself, and on the broader connected ecosystem.

To sustain that loop, background services run persistently. Action and attention are tracked. Resources are consumed. Updates reinforce the model. The system isn’t just operating at this point, it’s actively participating, on its own behalf.

The result isn’t a bug, so much as a business model.

The cost, too, isn’t just performance; it’s posture. Windows 11 no longer assumes the user is capable. It assumes you need steering, that deviation is synonymous with risk, and that user autonomy is noise to be managed, rather than intent to be respected.

The architectural plan this pattern exposes is one of containment, rather than trust… Which makes it fundamentally incompatible with serious work.

Engineering requires frictionless control; freedom to inspect, the ability to reconfigure, and the option to break what needs rebuilding. Writing requires stillness and focus, not distraction via suggestion notifications. Architectural integrity requires transparency: the ability to see how things fit, and to change them when they don’t. Creative and technical work assume the system is yours.

Windows 11 no longer behaves that way. It no longer treats the user as the owner of the experience, or even the customer that bought the license, but instead as the subject of it.

And to be clear: none of this is an accident, and none of it is illegal (in most places). Microsoft is entirely within its rights to design Windows around engagement. It owns the platform, and closed-source software is not a democracy.

But users still have one remaining vote: exit. Despite how it appears, leaving isn’t always an expression of defiance. Sometimes it’s just alignment, or a quiet refusal to stay in a system whose incentives no longer match your own.

The problem isn’t that Windows 11 fell short. It’s that it delivered… on goals I no longer share.

I’m not leaving because Windows is failing. I’m leaving because it is succeeding, just not at what I came here for.

---

If You’re Thinking About Switching

Leaving Windows doesn’t mean giving up functionality. In 2025, there are more approachable, polished alternatives than ever, including those that don’t require you to become a command-line power user or rewire your entire workflow from scratch.

If you’re looking to explore what’s out there, here are a few beginner-friendly Linux distributions (“distros”) that should feel familiar to long-time Windows users:

Linux Mint (Cinnamon Edition)

Best for: Simplicity and familiarity
Linux Mint is widely recommended for newcomers. Its Cinnamon desktop layout closely mirrors the traditional Windows experience: taskbar, Start menu, system tray—it’s all where you expect it. Mint emphasizes stability, ease of use, and a soft learning curve.

Kubuntu

Best for: Customization and aesthetics
Kubuntu pairs the Ubuntu base with the KDE Plasma desktop—one of the most modern and visually refined desktop environments available. If you want a system that feels polished and lets you tweak every detail without breaking things, Kubuntu is a great choice.

Bazzite

Best for: Gaming and out-of-the-box hardware support
Bazzite is a newer distro built on Fedora and optimized for gaming, media, and general desktop use. It ships with tools like Proton (via Steam) and Wine preconfigured, which makes running Windows games and apps far simpler than it used to be. It’s particularly good for users who want minimal setup and maximum compatibility.

Try Before You Commit

The beauty of Linux is that you can test-drive it without making any permanent changes to your system. Here’s a simple way to create a Live USB, which lets you boot into Linux temporarily and see how it feels:

What You’ll Need:

• A USB stick (8GB or larger)
• A free app called Balena Etcher (works on Windows, macOS, and Linux)
• The ISO file for the distro you want to try (download from the official site)

How to Try It Out:

1. Download the ISO file for Mint, Kubuntu, or Bazzite.
2. Open balenaEtcher, select the ISO, plug in your USB stick, and flash the image.
3. Reboot your computer and press the boot menu key (often F12, F2, or ESC depending on your device).
4. Select the USB drive and boot into the live environment.
5. Try the system without installing it, or click “Install” when you’re ready to make the change.

Running Windows Apps on Linux

One of the most common concerns is software compatibility, but modern Linux distros are more capable than you might expect. Many include built-in compatibility layers like:

• Wine – Lets you run many traditional Windows applications.
• Proton (via Steam) – Designed for gaming, but also works for general software.
• Bottles – A simple, GUI-based app for managing Windows software installs in Wine.

Between these and a growing number of native Linux apps (or web-based alternatives), most people can transition without losing access to their essential tools. And if you need Windows for a specific task, lightweight virtual machines or dual-boot setups remain viable options.

And, finally… The Linux of 2025 Isn’t What You Remember

If your impression of Linux comes from 2003-era forums and Bash-heavy tutorials, it’s time to take another look.

While the command line is still available (and powerful), most modern distros are GUI-first and user-friendly. You’ll find software stores, visual settings managers, update tools, and intuitive desktop environments. In fact, Linux doesn’t necessarily have more quirks than Windows these days, it just has different ones.

And perhaps most importantly, given the discussion above: you remain in control.

You choose what gets installed, what runs, what updates, and how much of your system is shared, monitored, or monetized. It’s not perfect, but it’s transparent, and in an era where control over your digital environment is increasingly hard to come by, I believe that matters.