Expand Without Drifting
A closer look at the first of the three AI-intensification tendencies — and why awareness, not judgment, is what separates expansion from drift.
Last week, in Are We Doing It to Ourselves?, I wrote about the three tendencies UC Berkeley researchers found in their eight-month study of AI at work: task expansion, blurred boundaries, and more multitasking. Workers at a 200-person tech company took on more tasks, worked longer hours, and ran more AI threads in parallel. Much of it wasn't asked of them — AI just made "more" feel possible, accessible, even rewarding.
This week, I want to ask something of you — about your own task expansion.
Have You Noticed Your Own?
Maybe you are drafting long, polished replies to emails that used to take three sentences.
Maybe you are building a small tool, a form, or a webpage yourself — something you would have hired someone to make.
Maybe you are researching a decision you would have made on instinct before.
Maybe you are writing more, publishing more, responding to more.
That’s task expansion. AI didn’t ask. It just quietly invited — and we said yes.
Three From My Own Week
I’ve been watching it in my own work, and it looks like this.
My books. I’m doing my own bookkeeping in QuickBooks now, with the help of ChatGPT. That used to be delegated work. Today, it’s part of my week. But my taxes? Those still go to my CPA. I want an expert to do that — someone with professional accountability and deep training I don’t have.
My writing. I’m writing this newsletter more consistently and producing better content. I’m also writing regularly on Medium and Substack — and the writing cadence has genuinely improved.
My session follow-ups. It is now part of my routine to send a summary email after every client session, using Fathom. Before, that was too time-consuming to add to my follow-up process. Now it takes me minutes — and my clients have a record of every session in their inbox.
All task expansion. All things I wasn’t doing two years ago — ages ago in AI time.
What Makes the Difference
Task expansion, by itself, is not good or bad. What makes the difference is awareness.
Doing my books with ChatGPT serves me — as long as taxes still go to my CPA. Writing more serves me — as long as the writing stays mine, not a shortcut that hollows out what I actually want to say. Sending session summaries serves my clients — as long as it supplements the relationship instead of standing in for it.
Awareness is what keeps expansion from drifting.
Each of my three examples has a line where it works. Cross the line and the same task expansion that was serving me starts to cost me something — my expertise boundary (taxes), my voice (writing), or my relationship depth (session summaries).
I only know where that line is because I’m paying attention.
The Pause Between
The pause between is how you find out — the moment of reflection that lets you see what an experiment is actually doing for you.
Not an afternoon retreat. Not a three-day journaling practice. Just the pause. The moment where you stop, notice, and ask: is this still serving what I’m actually here to do?
Without it, task expansion becomes momentum — which looks like progress from the outside and like drift from the inside.
The Practice
This is what my P.I.E. Method is built for:
Pause. Create the moment where you can see what’s actually happening.
Increase awareness. Notice what the experiment is producing — in your output, in your time, in your presence with the people who matter.
Embrace experimentation. Then reshape, redirect, or keep going — without guilt, without dogma.
You can’t know upfront whether a given task expansion is serving you. You can only know through living it, then stopping to reflect, and deciding on the next experiment.
That’s not a technology move. That’s a wisdom move.
One Question for the Week
So — what’s the task you are doing this week because AI invited you into it? And is it serving what you are actually here to do?
If you don’t know yet, that’s data too.
References:
Melikian, A., Are We Doing It to Ourselves?, KIT Newsletter, April 15, 2026 — https://open.substack.com/pub/anamelikian/p/are-we-doing-it-to-ourselves
Ranganathan, A. & Ye, X.M., AI Doesn’t Reduce Work — It Intensifies It, Harvard Business Review, February 9, 2026 — https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it