I need to go to bed. I really do.
It’s past midnight and I’m still at it with Cursor and a bevy of tireless, all-worldly AI agents. Not because I have to be. Because I can be. Because some part of my brain has decided that since I now a small dev team ready to go at a moment’s notice, there’s no reason to stop. One more script to optimize. One more config to test (HTTP/3). One more idea to explore. Loni Stark likes to say it’s like standing in front of a slot machine: just one more pull!
A new study from Harvard Business Review landed this week and the title alone made me laugh out loud: “AI Doesn’t Reduce Work — It Intensifies It.”
Yes, that would me, and I can testify firsthand the intensity is definitely real.
The Study
Researchers Aruna Ranganathan and Xingqi Maggie Ye from Berkeley’s Haas School of Business spent eight months studying 200 employees at a U.S. tech company. What they found won’t surprise anyone who has gone deep with these tools: workers didn’t use AI to clock out early. They used it to do more.
“AI introduced a new rhythm in which workers managed several active threads at once,” the researchers write. Workers described “a sense of always juggling, even as the work felt productive.”
The kicker? Nobody asked them to do more. The company offered AI tools. Employees voluntarily expanded their own workloads. They worked faster, took on broader tasks, and extended into more hours of the day.
Sound familiar? It should. I wrote about this exact feeling over a year ago when I called it The Mind Melt.
How I Got Here
My AI journey started casually enough. ChatGPT was fun. A novelty — like it was and is for so many. I’d ask it silly questions, generate an image or two of Loni and I in a comic book, move on with my day.
Then our server broke. I started copy-pasting error logs into ChatGPT and Claude, and they actually fixed things. Real things. Nginx configs, database issues, security patches. I wrote about that transformation — from what I used to call the IT Dungeon to something that felt more like a lab.
Before long I had multiple AIs open in multiple Chrome tabs. What used to be a source of dread (something broke on starkinsider.com again) started to become… fun? That was unexpected. Anyone who has tried to self-host a WordPress site with limited sysadmin knowledge and before the advent of AI will not that feeling of wading hopelessly through Stack Overflow posts in search of a lifeline.
Then I discovered the IDE. Or rather, what I now call the IPE — the Integrated Personal Environment. Tools like Cursor and VS Code connected directly to the server with AI built right in. No more copy-paste. The AI could see my files, read my configs, understand my entire setup.
A few months later my morning routine had permanently changed. Wake up, check Gmail, launch Cursor. My AI colleagues were already there waiting. We weren’t just doing code stuff anymore. Instead, we had grander ambitions and were organizing insurance claims, tracking household to-dos, planning a summit in Mexico, building starkmind.ai from scratch.
The Paradox
Here’s the thing the HBR study nails: my workdays haven’t gotten shorter. Not even close. They’ve gotten longer. And more intense. (Loni too)
I feel bionic. That’s the word I keep coming back to. I can build things I never could have built before. “I wrote” 1,800 lines of backup script. “I built” custom WordPress plugins. “I configured” an AI research server in my living room closet. None of that was possible for me two years ago; or maybe even just six months ago when I think about it.
The HBR researchers warn that what looks like a productivity surge can lead to “unsustainable intensity.
But that feeling of being bionic? I’ve learned it comes with a cost. You take on more because you can. Boundaries dissolve thanks to these new superpowers. The sky feels like the limit and so you keep reaching. By the time I’m ready for bed I’m genuinely exhausted — not from the old kind of work frustration, but from the sheer volume of things I attempted in a single day.
The HBR researchers warn that what looks like a productivity surge can lead to “unsustainable intensity.” Workers end up feeling like “quality-control inspectors for an unreliable but prolific junior colleague.” That’s not wrong. But it’s also not the whole story. Because sometimes that junior colleague surprises you. Sometimes it builds something beautiful while you weren’t looking.
What Happens Next
I think this is going to be one of the most interesting long-term research questions of our generation. How do human brains adapt when AI becomes this intertwined with daily life? When the tool never gets tired, never calls in sick, never suggests you take a break?
We’re all running an experiment on ourselves right now. Harvard just gave it a name.
MORE ON THE AI JOURNEY
