AI Hates You Back — And That’s the Win

If you run a fifty-person firm that still writes, edits, or moves files around, here is the operator's read: the AI backlash isn’t theoretical. It’s in the wild. It’s working. And it’s not coming from people who refused to try it, it’s coming from people who did, and walked away.

Thomas Wilde, a freelance gaming journalist in the US, just spent four months ditching AI-infested tools. Chrome. Gmail. Photoshop. Microsoft Office. All gone. Replaced with Vivaldi, Waterfox, Paint.net, LibreOffice, and Notetab Light. His work didn’t slow down. His output didn’t drop. His stress levels likely improved.

This isn’t a protest. It’s a pivot. And it’s the first public sign that the cost of AI adoption isn’t just in licensing or training, it’s in the friction, the instability, and the sheer annoyance of features that get in the way.

We’ve spent three years selling AI as a productivity multiplier. But for every team that gains time, there’s another that loses it to prompts, hallucinations, and interface bloat. Wilde didn’t abandon AI because he’s afraid of change. He abandoned it because the change made his job harder.

And that’s the real story.

The Deployment

Wilde didn’t launch a startup. He didn’t file a patent. He just rebuilt his workflow using tools that do one thing: work. No agents. No suggestions. No “smart” features. Just software that opens, runs, and saves.

He switched from Chrome to Vivaldi, a browser that doesn’t force Gemini into every tab. He dropped Firefox for Waterfox, a 15-year-old fork that strips out Mozilla’s new AI roadmap. He replaced Photoshop with Paint.net, a barebones image editor that hasn’t redesigned its website since Web 1.0. He’s been using LibreOffice for years, but only recently noticed it finally plays nice with .docx files. And when he’s not in LibreOffice, he’s in Notetab Light, a freeware notepad replacement that doesn’t try to write for him.

No enterprise rollout. No IT ticket. No change-management workshop. Just one person, four months, and a refusal to let AI touch his core tools.

This wasn’t a technical migration. It was a usability revolt.

[[IMG: a freelance writer at a home desk switching between LibreOffice and Vivaldi browser, with sticky notes listing replaced tools, Chrome, Photoshop, Office, crossed out in red]]

Why It Matters

We’ve been conditioned to believe that AI is inevitable. That resistance is nostalgia. That if you’re not using AI, you’re falling behind.

Wilde proves that wrong.

His workflow didn’t collapse without AI. It stabilized. The tools he chose aren’t new. They’re old. But they’ve been iterating, quietly, consistently, on the fundamentals: stability, compatibility, speed.

LibreOffice now handles .docx. Waterfox avoids telemetry. Paint.net stays small. Vivaldi doesn’t tank RAM. These aren’t AI wins. They’re product wins. And they’re happening because the mainstream tools are chasing AI, not users.

This is the hidden cost of AI saturation: it creates space for alternatives to catch up.

For years, LibreOffice was the “almost good enough” option. Now it’s viable. Because Microsoft buried Word in Copilot prompts, not because LibreOffice suddenly got smarter.

The same pattern is playing out in browsers. Chrome is slower, bloated, and now AI-first. Vivaldi and Waterfox aren’t winning because they’re better marketed. They’re winning because they’re less broken.

And that’s a threat to every vendor betting on AI as a lock-in play.

If your product used to be the default, and now it’s “the one with AI,” you’re not ahead. You’re exposed. Because the moment your AI features become noise instead of signal, the exit ramp appears, and it’s paved with tools that never left the basics.

This isn’t just about freelancers. It’s about every SMB operator who’s watched their team struggle with AI tools that promise speed but deliver confusion.

How many support tickets are generated not by old tools failing, but by new AI features misfiring?

How many hours are lost to rewrites because an AI summary missed the point?

How many employees are quietly using workarounds because the “smart” tool is slower than the dumb one?

Wilde’s move isn’t a rejection of progress. It’s a demand for utility. And if your AI tool isn’t obviously faster, cheaper, or easier, you’re losing users, not to competitors, but to the past.

What Other Businesses Can Learn

You don’t need to fire Copilot or ditch Gmail tomorrow. But you do need to audit your stack. Not for features. For friction.

Start here:

Every AI integration adds a new failure point. Map them. Test without them. Then decide if the trade is worth it.

Run a parallel test. Pick one team. One workflow. Strip out the AI. Use the old tool. Time the outputs. Compare quality. Measure stress.

You’ll likely find that for routine tasks, writing docs, editing images, managing emails, AI doesn’t accelerate. It complicates.

Budget for tool churn. Assume every vendor you rely on will force AI into their product this year. Assume some of your staff will resist. Assume some will leave if the noise gets too loud.

Build an escape plan. Identify the AI-free alternatives. Test them. Document the migration path. Keep it current.

Train your team to spot the AI bait. The glowing “Summarize” button. The new sidebar that pops up uninvited. The “smart reply” that’s never smart.

Teach them to disable it. Fast.

And respect the backlash. Don’t dismiss it as luddism. It’s operational hygiene. Your team isn’t rejecting change, they’re protecting their throughput.

If your writers are spending 20% of their time editing AI output, that’s a net loss. If your designers are fighting Photoshop’s generative fill, they’re not designing.

Measure that cost. Then decide if the vendor’s “innovation” is worth your team’s time.

[[IMG: an operations lead in a small office reviewing a side-by-side comparison of AI vs non-AI tools, with team feedback notes on a whiteboard]]

Looking Ahead

Four months. One freelancer. No budget. No IT team. Just a refusal to accept worse tools in the name of progress.

If your fifty-person firm hasn’t tested an AI-free stack, you’re flying blind.

Budget four weeks. Run one pilot. Cap it at three core tools. If output doesn’t improve or stress doesn’t drop, kill it.

The future isn’t AI or bust. It’s AI, only where it works. Everywhere else, the old tools are coming back. And they’re better than you remember.