Your AI 'Assistant' Isn't Your Helper. It's Your New Boss.

I fell for it, just like everyone else. A new suite of AI tools landed in my company’s software stack, and the pitch was seductive. "Your partner in productivity." "A new way of working." "Your own personal AI assistant."

I’m a writer. My work is a messy, unpredictable stew of caffeine, panic, and the occasional flash of inspiration. The idea of an "assistant" to handle the boring bits—summarizing long documents, transcribing interviews, drafting boilerplate emails—sounded great. So I switched it on.

For a week, it felt like magic. And then, a creeping dread set in. The "assistant" wasn't just assisting. It was… watching. Nudging. Suggesting. It proposed more "efficient" ways to structure my documents. It analyzed the sentiment of my draft emails. It helpfully plotted out my project timelines based on my recent activity.

My AI assistant wasn’t a partner. It was a manager. A micromanager, to be precise. And this isn’t a bug; it’s the entire point.

The Lie of "Assistance"

Let's be clear: the biggest names in tech, from Microsoft with its Co-pilot to Google with its Workspace AI, aren't selling you an assistant. They're selling your boss a new form of surveillance. They use friendly, anthropomorphic language—"co-pilot," "helper," "buddy"—to mask the cold, mechanical reality.

These tools don’t help you do your work; they integrate themselves into your workflow. They become the inescapable interface through which your work is performed. Every query you make, every document you start, every "suggestion" you accept or reject is a data point.

This isn't just improving the product. It’s building a detailed, minute-by-minute performance profile of you.

It’s the digital equivalent of a manager standing over your shoulder, stopwatch in hand, except this manager has a perfect memory and zero empathy. It knows how long you spent on that report, how many versions you drafted, and how your "productivity" this Tuesday compares to last Tuesday.

We’ve seen this playbook before. Algorithmic management has been squeezing gig workers, delivery drivers, and warehouse staff for years, turning human beings into cogs in a pitiless machine. Their every move is tracked, their "efficiency" is constantly scored, and they can be deactivated by an algorithm for reasons they'll never understand. We, the comfortable white-collar professionals, looked on with a vague sense of pity, never imagining it would come for us. Well, it’s here. It just showed up in a nicer user interface.

The Slow Death of Professional Judgment

The real danger isn't just the surveillance. It's the slow, insidious erosion of our own autonomy and skill.

When the AI is always there, "assisting," you start to defer to it. Why wrestle with a difficult first draft when the AI can spit one out in seconds? Why struggle to find the right words for a sensitive email when the AI can offer five blandly corporate options? Why use your own hard-won experience to map out a project when the AI can generate a "best-practice" timeline instantly?

Each time you accept the AI’s suggestion, you give away a small piece of your professional agency. You’re no longer the expert; you’re the human-in-the-loop, the fleshy validator for the machine’s output. The AI isn't your assistant. You are its.

This process of de-skilling is terrifying. Our professional judgment is a muscle, built over years of making mistakes, trying things, and developing our own unique approach. These AI "assistants" are designed to let that muscle atrophy. They steer us toward the mean, the average, the "most likely" response. They promise to kill writer's block but really just kill the incentive to think for yourself.

And you can't even argue with your new boss. You can’t tell an algorithm that its "optimal" solution is actually terrible for a long-term client relationship. You can’t explain to a language model that the most efficient path isn't always the most creative or humane one. You are managed by a black box you cannot question. It’s a recipe for a new, profound kind of workplace anxiety.

Unplug from the Machine

We’re being sold a false bargain: trade your autonomy and professional expertise for a bit of convenience. It’s a bad deal.

We have to consciously reject this future. We have to carve out spaces for un-surveilled, un-quantified, truly human work. This isn’t about being a Luddite; it’s about being a professional who values their own mind. Turn off the "assistance." Close the chatbot. Insist on the messiness and inefficiency of genuine human thought.

Your best ideas will not be "suggested" to you by an algorithm. They will come from the quiet, analog spaces you protect from the ever-watching eye of your new AI manager.

***

Analog Recommendations

If you're ready to reclaim your focus and privacy, here are two tools for unplugged work: