The AI Productivity Trap: Why Automating Everything Is Burning Us Out

 

Created with DALL-E prompting by author

---

Remember when we thought AI would set us free? I do. It was 2024, and I sat in a conference room listening to a consultant pitch on "AI-powered workflow optimisation." The promise was intoxicating: automate the mundane, elevate the meaningful. No more tedious email sorting. No more mind-numbing data entry. No more transcribing hour-long meetings by hand.

We'd finally have time to think

A recent study featured in the Harvard Business Review extends that premise to its logical conclusion, revealing that rather than a productivity revolution, companies face the danger of evolving into burnout-inducing environments.
As part of what they call “ongoing research,” UC Berkeley researchers spent eight months inside a tech company with 200 employees, observing what happened when workers truly adopted AI. From over 40 “in-depth” interviews, they discovered that no one at the company felt pressured or was asked to meet new targets. Instead, people simply started accomplishing more because the tools made more tasks seem achievable. However, as they could do more, work began to extend into lunch breaks and late evenings. The employees’ task lists grew to fill every hour that AI freed up, and then continued beyond that.

The Promise That Broke Our Brains

The initial promise of Artificial Intelligence was straightforward: it would handle the tedious tasks. Bots(agentic AI) would manage our emails, process expense reports, and transcribe meetings. In return, we would gain freedom to focus on big ideas, strategies, and innovation. It sounds ideal. For many professionals in 2026, it feels more like a mental cage.

Here's what nobody told us: removing all the low-effort "busywork" from our day doesn't grant freedom. Instead, it leaves us with 8 to 10 hours of intense, nonstop critical thinking. And here's the uncomfortable truth - the human brain isn't built to sustain a sprint-like marathon.

Think about your typical workday before AI assistance became ubiquitous. You'd spend maybe two hours on genuinely challenging cognitive work - architecting a new system, solving a complex design problem, strategising a client pitch. The rest? Emails. Social Media, Status updates. Formatting documents. Administrative tasks that felt like friction. But here's what we didn't realise: that friction was actually functional.

The Neuroscience We Ignored

Your brain operates in two fundamental modes, and understanding this changes everything. There's focused mode - the intense, analytical thinking you do when debugging code, designing a user interface, or architecting a database. It's cognitively expensive. It burns glucose. It requires your prefrontal cortex to fire on all cylinders. Then there's diffuse mode - the relaxed, wandering state your brain enters when you're doing something routine like responding to simple emails and organising files and yes, even filling out timesheets. Neuroscientists have known for decades that breakthroughs don't happen during focused thinking. They happen during diffuse mode. That's when your brain makes unexpected connections. When solutions to problems you weren't actively thinking about suddenly crystallise.

You know that moment in the shower when you suddenly solve a coding problem that stumped you all afternoon? That's diffuse mode at work. By automating away every routine task, we've inadvertently eliminated our brain's recovery time. We've removed the very conditions that foster creativity, insight, and innovation - the things AI was supposed to help us achieve.

The Developer Who Couldn't Code Anymore

There is this developer called Marcus, a developer I've known for four years. He was early to adopt AI coding assistants - Cursor, GitHub Co-pilot, the works. His productivity metrics skyrocketed. He was shouting from the rooftop about shipping features at twice his previous pace.

Six months in, something changed.
"I can't think in code any longer," said Marcus. "I describe what I want, the AI writes it, I review and deploy. But when I sit down to actually design something new, something the AI can't just generate from a prompt... my mind goes blank." Marcus had outsourced so much of his actual coding to AI that he had atrophied his core problem-solving skills. The routine task of writing basic functions - something that used to feel like busywork - was actually keeping his architectural thinking sharp. He's not alone. This pattern repeats across disciplines. Designers who can no longer sketch rough concepts because AI generates polished mock-ups instantly. Consultants who struggle to structure presentations because AI assembles slide decks automatically. Writers who have forgotten how to develop ideas because AI completes their sentences.

The busywork wasn't just busywork. It was practice. It was a warm-up. It was the cognitive equivalent of a musician practising scales.

The Illusion of Infinite Capacity

Here's the insidious part: when AI handles all your routine tasks, your calendar suddenly looks wide open. And in our achievement-obsessed culture, empty calendar space feels like an invitation - or an accusation. So, what do we do? We fill it. More meetings. More strategic initiatives. More high-level thinking. After all, if AI freed up four hours of your day, you should use those four hours productively, right?

Wrong.
This is where the mental burnout accelerates. You're now spending your entire workday in focused mode, with no natural breaks, no mental recovery periods, no time for your brain to wander and make connections.

A designer I know - let's call her Sarah - described it perfectly: "Before AI tools, my day had a rhythm. Deep work, then some email. Creative thinking, then some file organisation. Hard stuff, then easy stuff. Now it's just hard stuff, back-to-back, all day long. By 2 PM, I'm cognitively cooked, but I've still got six hours of 'focus time' blocked on my calendar." The cruel irony? Sarah is less creative now than she was when half her day was "wasted" on mundane tasks.

What We Lost in the Optimisation

There's a concept in music called "negative space" - the silence between notes that gives music its emotional impact. Remove the silence, and you don't get more music. You get noise.

Our work lives are the same way.

Those "inefficient" moments - waiting for a file to upload, walking to a colleague's desk instead of texting them, reading another email from the Nigerian Prince with money to burn (scammer), manually formatting a document - weren't pure waste. They were negative space. Moments where our conscious mind rested, and our subconscious got to work. A UX researcher who made a deliberate choice to stop using AI transcription for user interviews. "The AI transcripts were perfect," she explained. "But when I transcribed manually, I'd notice things. A participant's hesitation. A phrase they repeated. Patterns I'd completely miss if I just read the AI summary."

The inefficiency was the insight.

The Paradox of Hyper-Productivity

We're creating a generation of knowledge workers who are simultaneously more productive and less capable. More productive because AI handles execution brilliantly. Less capable because we're losing the muscle memory of actually doing the work. A junior developer recently asked me how to debug a complex async issue. Not how to prompt an AI to fix - she knew that. She wanted to understand the underlying problem. And I realised: if I'd started my career with AI assistants handling all the "tedious" debugging, would I have the deep systems knowledge I rely on today?

Probably not.

We're optimising for short-term output at the expense of long-term capability. And the scary part? We won't realise what we've lost until we need skills we never developed.

Breaking Free From the Productivity Prison

So, what do we do? We can't un-invent AI, and frankly, I wouldn't want to. But we need to be smarter about how we integrate it into our work.
First, we need to reframe "busywork." Not all routine tasks are worth automating. Some are worth keeping specifically because they give your brain recovery time. One developer I know still writes his commit messages manually, even though AI could generate them. "It's my moment to reflect on what I actually accomplished," he says. Second, we need to design deliberate inefficiency into our days. Take walks between meetings. Sketch ideas by hand before jumping into design tools. Write rough drafts without AI assistance. These aren't productivity killers - they're creativity enablers. Third, we need to resist the tyranny of the optimised calendar. Just because AI freed up time doesn't mean you should immediately fill it with more cognitively demanding work. Guard that time. Use it for exploration, learning, or genuine rest. Fourth, maintain your fundamental skills. Even if AI can do it faster, occasionally do it yourself. Code without autocomplete. Design without AI mock-ups. Write without sentence completion. Think of it like a professional athlete who still practices basic drills despite being elite.

The Human in the Loop

Here's what I've come to believe: the future of work isn't about humans or AI. It's about humans with AI, used thoughtfully. The best developers I know use AI coding assistants, but they're selective. They automate the truly mundane - boilerplate code, basic CRUD operations - but they hand-craft the architecturally significant code themselves. The best designers use AI for rapid iteration, but they still sketch concepts manually first. The AI enhances their vision; it doesn't replace their thinking. The best consultants use AI to analyse data and generate initial frameworks, but they still do the hard cognitive work of synthesis and strategic insight themselves. They've learned something crucial: AI is a tool, not a replacement for thinking. And like any tool, its value depends entirely on how you use it.

Reclaiming Our Cognitive Freedom

The promise of AI was freedom. Freedom from drudgery. Freedom to focus on meaningful work. But we forgot that meaning comes from the entire experience of work - not just the highlights. The sculptor doesn't just envision the statue; they chip away stone, brush off dust, and smooth rough edges. The process matters.
This week, identify one "inefficient" task you've automated. Do it manually. Write those commit messages by hand. Transcribe that meeting yourself. Organise those files without AI assistance.

Notice how it feels. Notice what your mind does during those moments. You might be surprised to find that the freedom you've been seeking doesn't come from eliminating every mundane task. It comes from having the space to think, to wander, to be human. The AI revolution promised to free our minds. Maybe it's time we used that freedom not to think harder, but to think better. And sometimes, that means embracing the tedious, the routine, the supposedly inefficient moments that make us more than just biological task managers.

Because at the end of the day, the goal isn't to become more machine-like in our efficiency. It's to become more human in our creativity, wisdom, and insight. And that requires giving our brains what they actually need: not endless optimisation, but rhythm, rest, and the occasional mundane task that lets our subconscious work its magic.