The Problems I Solve Are Never Solved at the Desk

There is a particular kind of stuck that a screen cannot fix.

It shows up two or three days into a hard problem — usually an architectural one, where there is no bug to find and no test to make pass, just a decision with four plausible answers and consequences that won't reveal themselves for another two years. I can sit with it for eight hours and produce nothing but four slightly better-argued versions of the same indecision.

The fix, for me, is to get on a motorcycle and leave Jaipur.

Why it works, as far as I can tell

I don't think there's anything mystical in it. Riding occupies exactly the part of your attention that would otherwise be generating noise. You have to watch the road, so you cannot rehearse the same three arguments in a loop. What's left over gets to work on the actual problem, quietly, without supervision.

The other thing is that you can't act on anything. No editor, no branch, no way to start implementing the first idea that sounds reasonable. Being unable to build the wrong thing for six hours turns out to be a feature. Most bad architecture I've seen — including mine — came from being able to start immediately.

What Rajasthan gives you

Distance, mostly, and not much interruption. The roads out of Jaipur will give you a couple of hundred kilometres of open country if you want them, and the landscape is repetitive in a way I've come to appreciate — there isn't enough novelty to hold your attention hostage. Aravalli hills, scrub, a town, more scrub.

It's the opposite of a holiday. A holiday demands that you notice things. This demands nothing at all, which is the point.

The honest version

I'd like to claim this is a deliberate practice. It isn't, really. It started as wanting to be away from work and turned into the most reliable debugging tool I have, and I only noticed the pattern years later — that a disproportionate number of the decisions I still stand behind were made somewhere between two towns, and written down when I got back.

The writing-down part matters. Whatever you worked out on the road does not survive the first hour of email. It has to go into a document before anything else happens.

That's the whole method. Ride until the noise stops, then write it down before it comes back.