【Text】
I’ve been staring at my todo list for the past week, and something finally clicked.
Not about the list itself. We all know that feeling—the list that grows faster than you can check things off. The real click was about the stuff on the list. Which items made me feel alive, and which ones made me want to close the laptop and go for a walk?
There’s this weird pattern I’ve noticed in my own work habits. The things I procrastinate on the most? They’re almost never the hard creative problems. They’re the tedious, repetitive, mind-numbing tasks that have to get done but add zero joy. Formatting documents. Sorting emails into folders. Writing boilerplate code for the hundredth time. Summarizing meeting notes that nobody will read anyway.
And the crazy part? I used to feel proud of grinding through those tasks. Like, look at me, I’m being productive. I’m getting things done. That’s a trap—the feeling of busyness masquerading as progress.
I started experimenting with this idea about six months ago. Every morning, I’d look at my day’s work and ask one question: “If I had an intern who was infinitely patient, never complained, and worked 24/7, which tasks would I hand off to them?” Then I’d take those tasks and figure out how to automate them with AI.
Turns out, the list was way longer than I expected.
Writing standard email responses? Hand it over to a custom GPT that knows my tone. Researching basic facts about an industry I’m not familiar with? Let Claude search the web and summarize. Drafting the first version of a project brief? Give the AI the bullets and let it turn them into paragraphs. Proofreading my own writing for typos and awkward phrasing? Please, take it all.
Here’s what nobody tells you about this approach. It’s not really about saving time. Not primarily, anyway.
What actually happens is more subtle. When you strip away all the tasks you could automate, what remains is the work you actually want to do. The work that requires your specific taste. Your judgment. Your weird, personal way of seeing the world.
I’m not going to lie and say this is easy. The hardest part isn’t even setting up the AI workflows. It’s the emotional shift. You have to admit that some of the work you’ve been doing—maybe the work you’ve been proud of doing—is actually just busywork that a language model could do in seconds.
That stings a little. It did for me, anyway.
But once you get past that, something liberating happens. You start to see the shape of your own contributions more clearly. The parts of your work that are genuinely you. The parts that require your personal history, your intuition, your ability to make a judgment call that doesn’t follow the rules.
Last week, I spent three hours writing a single section of a product design document. Three hours. For maybe 200 words. And I could feel the pressure mounting—the voice in my head saying, “You could have written 2000 words in that time using AI.” But I kept going. Because those 200 words required decisions that I couldn’t explain to an AI. They required me to think about what this product means to the people who use it, not just what it does.
That’s the stuff worth keeping.
Let’s be honest about something uncomfortable. A lot of the fear around AI replacing jobs isn’t really about AI being too smart. It’s about people realizing that a huge chunk of their daily work is the kind of thing that can be automated. The part that hurts is the implication that maybe that work wasn’t all that valuable in the first place.
But here’s the flip side. The parts of your work that require real thinking, real judgment, real taste? Those become more valuable. Not less. Because now you’re not competing on who can churn out the most boilerplate. You’re competing on who can see the thing that nobody else sees.
I’ve been running a quiet experiment with my team. I asked everyone to list the top five tasks they do that make them feel excited about their work. Then the top five that drain them. The results were depressing and exciting at the same time. Depressing because the “drain” list was full of stuff that could be automated yesterday. Exciting because the “excited” list was full of stuff that AI can’t touch yet.
The pattern was clear. The draining tasks were all about reproducing existing patterns. The exciting tasks were about creating new ones.
This is where the whole “AI will take our jobs” conversation gets it backwards. AI isn’t coming for the jobs that require creativity and judgment. It’s coming for the jobs that require sitting in an office copying data from column A to column B and pretending that’s value.
The real challenge isn’t learning how to use AI. That part is easy. The real challenge is learning how to stop doing the work that makes you feel busy but doesn’t move the needle. The real challenge is having the courage to say, “I’m going to trust the machine with the boring stuff, and I’m going to spend my limited attention on the stuff that matters.”
Every week, I look back at what I actually did. Not what I was busy doing. And the weeks where I pushed more of the “don’t want to” work to AI? Those were the weeks where something interesting happened. Where I had a new idea. Where I saw a pattern nobody else saw. Where the work felt like mine.
That’s the trade worth making.