The Scene That Made Me Write This
It's Sunday. I'm supposed to be resting.
Instead, I have Claude Code running in two terminal windows, debugging two different apps simultaneously. In another tab, I'm drafting this blog post. My to-do list for today has twelve items on it, and somehow that feels... manageable? Because with AI, twelve items isn't what it used to be.
This is the moment I realized: something has to change.
Not because the productivity isn't real. It's very real. But because I'm running on fumes and I can't seem to stop.
The Acceleration Nobody Warned Me About
Here's what AI productivity actually looks like:
Tasks that used to take weeks? Minutes.
Projects that used to take months? Weeks.
Ideas that used to stay in my "someday" folder? Built and deployed.
In the past ten weeks, I've created more value than I did in the past couple of years. That's not hyperbole. That's my actual output compared side by side.
And it's absolutely incredible. I'm riding the AI wave, and the view from up here is extraordinary.
But here's what nobody mentions in the productivity guru threads: when you can build this fast, it's almost impossible to stop.
The December Sprint I Didn't Plan
I was supposed to rest.
The last week of December, I took time off from work. I had plans: slow mornings, maybe some reading, definitely not staring at code.
Then Anthropic doubled our tokens for the week.
Free credits. Limited time. Use them or lose them.
(If you've read my earlier posts, you know I have a complicated relationship with Anthropic credits. This felt like déjà vu from November.)
So instead of resting, I did what any reasonable person would do: I generated twenty ideas for productivity apps. Then I started building them. All of them.
What Twenty App Ideas Looks Like
Here's a taste of what came out of that sprint:
Meetonomix — A meeting cost calculator. Because someone should know how much that "quick sync" actually costs when you multiply hourly rates by attendees by time.
MeetingQuant — An app for logging decisions made during meetings so they don't vanish into the ether. (How many times have you left a meeting and thought, "Wait, what did we actually decide?")
Bragging App — For knowledge workers who go through performance reviews. Track your accomplishments all year so when review season comes, you're not scrambling to remember what you did in March.
Focus Session Logger — A local productivity app that helps you focus. This one's already published and free. Anyone can grab it from my website.
Career Growth Tracker — Helps employees plan their advancement: what skills they need, what goals to hit, what conversations to have.
Interview Prep Tracker — For when you're job hunting and need to keep track of companies, prep notes, and follow-ups.
And that's just six of the twenty.
Each project is at MVP stage. Some will be free to try. I'll be giving access to these apps in the next few weeks to anyone who wants to test them.
The Cost of Creation
Let me be honest about the other side of this equation.
My AI tool spend increased 30% this month.
I was already paying for ChatGPT and Claude Code with a Max subscription. That felt expensive but justifiable.
Then I added:
- Whisper Flow — for voice transcription
- Google Gemini — for another perspective on problems
- Google AI Studio — for API access
- Notebook LM — for research and synthesis
Each tool is incredible. Together, they make productivity absolutely explosive. But the costs add up.
I'm now running multiple AI subscriptions, burning through tokens at a pace I couldn't have imagined six months ago, and producing more output than ever.
The ROI is there. I know it's there. But watching that number climb every month requires a new kind of budgeting conversation with myself.
The Exhaustion Nobody Posts About
Here's the part that doesn't make it into the "look what I built with AI" tweets:
I am tired.
Not the normal kind of tired. The kind where your brain has been running at 120% for weeks and it's starting to protest. The kind where you wake up with ideas and go to sleep with ideas and dream about code in between.
When progress is this fast, when results are this immediate, when the gap between "I wonder if I could..." and "it's deployed" shrinks to hours—it creates a strange new problem.
You can't stop.
Not because someone is making you. Because you don't want to. Because every hour you're not building feels like wasted potential. Because the tools are right there, waiting, and you know exactly what you could create if you just kept going.
This is the side of AI productivity nobody talks about. The pace is exhilarating and exhausting in equal measure.
Learning to Live at This Speed
I don't have this figured out yet.
What I know is that today—this Sunday where I'm debugging two apps while writing a blog post—something has to shift.
The productivity is real. The output is real. The value I'm creating is genuinely more than I've ever created before.
But sustainable? That's the question I'm sitting with.
In the next few weeks, I need to find a new balance. Not slower, necessarily. But smarter. With rest built in. With boundaries that actually hold.
Because the AI wave isn't slowing down. If anything, it's accelerating. And I want to ride it for years, not burn out in months.
What I'm Taking Away
Here's what I know to be true:
The productivity is not hype. I'm living proof that AI tools can compress timelines dramatically. Weeks to minutes. Months to weeks. It's real.
The cost is real too. Both financial (30% increase this month) and physical (my brain needs a vacation from itself).
The hardest part isn't building. It's stopping. It's choosing rest when progress is calling. It's closing the laptop when you know you could ship one more thing.
This will become a balancing act. For me, for you, for anyone who gets good at using these tools. The constraint is no longer capability. It's sustainability.
Your Turn
If you're feeling this too—the excitement and the exhaustion tangled together—know that you're not alone.
The pace is new for all of us. We're learning how to live with tools that make us dramatically more capable than we were twelve months ago. There's no playbook for this.
What I can offer: Some of the apps I built during this sprint will be free to try. Focus Session Logger is already available. More coming soon.
Maybe they'll help you be productive. Maybe they'll help you track your wins. Maybe they'll help you have better meetings.
And maybe, eventually, we'll all figure out how to ride this wave without drowning in it.
Still building, still exhausted, still trying to figure out where the off switch is—and whether I actually want to use it.
P.S. It's now 2pm on Sunday. I've finished this blog post. The two apps are still debugging in the background. My to-do list has eight items left.
I'm going to close my laptop now. I think.
P.P.S. If you want early access to any of the productivity apps I mentioned, keep an eye on my newsletter.