Top of the morning — I'm Janne Parkkila.
I work with AI for a living, and write about it the way I'd explain it to a friend.
No mystique, no doom, no hype. Just honest notes on where this technology genuinely helps, where it quietly doesn't, and what it's like to actually build with it.
02
Questions I keep coming back to
The handful of problems I find myself thinking about most — see all the questions →
Judging an AI honestly, without fooling ourselves
It's surprisingly easy to convince yourself a model is brilliant. I care about ways of testing that survive contact with the real world — measuring what it can truly do, not what a benchmark happens to reward.
When AI that “works” meets messy, human input
A system can look perfect in a demo and fall apart the moment a real person uses it. I spend a lot of time in that gap — the part the tutorials quietly skip.
Where extra complexity actually earns its keep
Sometimes a fancy, autonomous setup is the right call. Far more often, one clear instruction does the job. I like figuring out which is which before building the expensive thing.
04
A bit about me
I spend my days in the unglamorous middle of the AI conversation.
So much of the talk about AI lives at two extremes. It is either breathless excitement on one side, or flat dismissal on the other. I find both a little exhausting. The interesting part, to me, is the quiet middle: what this technology is genuinely good at, where it lets you down, and what it takes to put it in front of real people without everything falling over.
I wear two hats, and they keep each other honest. As a researcher, I read the papers, run the tests, and try to stay current with what actually changed this month (or yesterday!). As a developer, I build products and live with the consequences when they don't work (like software usually does). Neither half lets the other get away with hand-waving.
If you're a founder weighing a decision, a researcher comparing notes, or just someone trying to make sense of a noisy field, welcome! I write for you. Clear thinking about a genuinely complicated subject, without talking down to anyone. That's the whole job, really.
05
Say hello
I'd genuinely like to hear from you.
A good question, a paper worth reading, a project to compare notes on, or just a hello. I try to read everything that comes in. No pitch necessary, and no wrong reason to write. And please, no AI slop.