Obsession has a branding problem.

Say the word out loud and people picture something clinical: a guy rearranging his pens at 3am, a spreadsheet formula that’s quietly eaten someone’s summer, a disorder with its own page in a medical manual. The dictionary agrees. It files obsession right next to unhealthy. But the obsession I want to talk about is the one that’s carried me, quietly, into every room I’ve walked into not knowing a thing. And I’ve walked into a lot of those rooms.

Here’s the thing about obsession. It always leaves a tell.

You can’t fake it and you can’t hide it. It shows up in the work whether you wanted it to or not.

You can spot it from across the room

An obsessive carpenter does not leave a messy corner. Doesn’t matter that no one will ever open that cabinet and run a finger along the inside joint. He knows it’s there. The clean corner is the tell.

Zoom all the way out and the same thing holds. A civilisation obsessed with making its people’s lives easier can’t help but show it. It leaks into how the government organises itself, what it bothers to measure, which problems it has decided are simply unacceptable. You don’t need to read the mission statement. You can read the obsession straight off the institutions.

That’s the first useful thing about it: obsession is legible. It’s one of the few human qualities you can audit from the outside, without being told. (Try auditing “passion.” Good luck. 😛)

“Once you decide on your occupation you must immerse yourself in your work. You have to fall in love with your work… You must dedicate your life to mastering your skill.” (Jiro Ono, Jiro Dreams of Sushi)

At work

Every job I’ve ever taken has gone exactly like this. I walk in knowing nothing. Not false-modesty nothing. Actually nothing. Everyone in the room has more context than me, more reps, more scar tissue.

The only lever I have is that I can’t put the thing down. I obsess over the corner nobody’s checking. And that, not talent, not pedigree, buys me a minor edge. The edge buys a little confidence. The confidence buys slightly bigger problems to obsess over. And round it goes. Every single time.

That’s the quiet mechanism no one tells you about: obsession compounds. It is the only starting capital you can manufacture out of thin air when you have nothing else to bring to the table.

And then I pointed that loop at the biggest unknown I could find. I left a perfectly good job to build something from scratch. Not something that merely works, but something people would actually want to use, which is a far higher bar than it sounds.

Day 0, my cofounders and I knew nothing. The same way as before. Actually nothing. Not about voice. Not about LLMs. Not about how a GPU really works once you get past the marketing. Three people obsessing over a problem we had no business being in the room for.

What carried us wasn’t a head start. It was that we kept obsessing over details the market didn’t care about. Not early, anyway. The slice of latency nobody would consciously notice. The edge case that would show up once in ten thousand calls. The market catches up to those details eventually. By the time it does, you want to already be the people who’ve been obsessed over them for years.

That’s how a thing nobody asked us to build ended up solving critical outcomes for legendary companies at scale. It wasn’t strategy. It was the same loop (knew nothing, couldn’t put it down, obsessed over the corner no one was checking), just run at a bigger problem, with people who couldn’t put it down either.

And we’re nowhere near done. The building is still very much on. But here’s what I’m certain of, looking back: you could have handed us the best researchers and the sharpest people alive, and we’d still have built nothing without someone obsessing over the problem 24x7x365. The talent was never the missing piece. Obsession is the thing that turns it over.

“Greatness is not intelligence. Greatness comes from character. And character isn’t formed out of smart people, it’s formed out of people who suffered.” (Jensen Huang)

Why it actually works

Strip it down and obsession does two things.

First, it makes you break things open. You cannot obsess over a problem and leave it sitting there as a black box. You keep prying until you can see the gears. Do that for long enough and you stop standing at the edge of a subject; you walk all the way to its frontier, the place where the half-broken-open problems live. That’s where the interesting work is. Obsession is just the thing that walks you there.

Second, obsession leads to clarity. When you care about one thing far more than is reasonable, the noise around it goes quiet on its own. You stop choosing what matters. The obsession already chose. Decisions that paralyse everybody else turn trivial, because there’s really only one variable you’re solving for.

A rule I’ve started trusting:

“if a decision feels hard, you’re probably not obsessed enough about any one thing inside it yet.” – SST, 2026

Especially now

We live in possibly the worst time in history to be un-obsessed.

Everything looks worth doing. Every skill has a 30-day roadmap, every side quest has a creator swearing it changed their life, every tab you open is a perfectly reasonable thing to give a year to. A thousand open tabs, all of them green, none of them really yours.

Obsession is what closes the tabs. Not by willpower. By gravity. It pulls you down into one thing hard enough that the other ninety-nine stop tugging at your sleeve. In a world engineered to keep everything looking equally shiny, that gravity is a quiet superpower.

AI made it harder, not easier

Here’s the cruel little irony of this moment. AI was supposed to lower the bar, and it did. It has never been cheaper to start. A landing page in an afternoon. A piecemeal product by the weekend. Code that used to need a team now falls out of a prompt. The floor went through the roof.

But the floor was never the hard part. The last mile was. And the last mile is exactly the bit AI doesn’t do for you.

Everyone now says coding is easy. Sure. What comes out the other end is a rough-around-the-edges product. And just like in real life, nobody wants to buy the rough-around-the-edges anything. The gap between “it works” and “I’d actually pay for this” is the same gap it has always been. It’s only more crowded now, because everybody cleared the floor at once.

That gap has a name. It’s taste. And here’s the part the tooling cannot shortcut: taste only comes from obsession. You don’t download it from a model. You earn it by caring about the corner nobody’s checking, over and over, until you can feel when something is off before you can explain why. AI will happily write the code. It cannot make the code have taste. That part is still, stubbornly, on you.

“The only problem with Microsoft is they just have no taste. They have absolutely no taste.” (Steve Jobs)

So

It’s the only reliable road I know to exceptional outcomes. Exceptional, by definition, is not what shows up when you spread yourself evenly across everything on offer. It’s what shows up when you go absurdly, disproportionately deep into one corner of the world, and leave a tell so loud that people mistake it for genius.

So obsess. Pick the corner. The world is strangely generous to the people who can’t stop.

This isn’t a guide on what to obsess over. That part’s yours, and I’m still figuring out mine 🙂

Also,Welcome to the corner after a long long time!

General