Watch a few episodes of Anthony Bourdain’s Parts Unknown, and you’ll notice a pattern: the master chefs care more about the ingredients than the recipes.
Masa Takayama, the legendary Japanese sushi chef, spent years apprenticing in Tokyo. The first two were spent wiping toilets, washing dishes, and cleaning fish. Only in his third year was he trusted with the rice. Eventually, he sliced fish and prepared nigiri. He respected the rice and fish so much that he was willing to spend seven years learning how to combine them well.
Another sushi master, Naomichi Yasuda, has spent nearly three decades running one of New York’s best sushi bars. Yet he still has a reverence for simple, quality ingredients. When Bourdain asks what the most important part of his cooking is, he says the rice.
BOURDAIN: Which is more important, the rice or the fish? Or what percentage? Rice. More important?
YASUDA: About 90 percent.
BOURDAIN: Wow.
YASUDA: Fish is the second ingredient. The main ingredient is rice. So my sushi is rice.
Even Masa’s mother understands the primacy of ingredients. When Bourdain visits, she sets out a pile of uni and fatty tuna brought up from that morning. She serves it plain, without attempting to cut it as masterfully as her son might. “This is better than Masa’s sushi,” she jokes. Masa smiles and concedes that she might be right.
We consumers instinctively value the ingredients more than the chef, too. It’s why sushi itself is internationally beloved; #sushi was the most popular dish on IG in 2021, appearing in over 3.2 million posts. A well-prepared, two-ingredient dish satisfies eyes and bellies better than a complicated one that tastes as if a committee made it.
Good ingredients are enough. The chef’s job is to let them be.
If software engineers are digital chefs, what are our ingredients?
Data.
All we do is show users data and help them change it. It’s not about algorithms, architecture, models, features, or P90 metrics. It’s the data, silly.
Rich Hickey understood this. He became so sick of complex programs that he quit his job and created Clojure. Instead of tempting devs to obsess about class hierarchies or frameworks, the language helps them work with the data directly by limiting itself to a handful of simple structures: maps, vectors, lists, and sets. Rich proved how powerful this simple language could be by building a successful database company upon it. A few basic building blocks, masterfully combined. Rice and fish.
I think about how beautiful Rich’s old approach is every time a new startup sends me a changelog that reads like a Walgreens receipt. New dashboard! New integrations! New collaboration features! New mobile app! New settings for the settings!
They’re shipping, all right. But for whom?
Nobody opens a calendar to admire its architecture. They want to see what’s happening this week. Nobody opens a notes app to celebrate its block editor. They want to remember something. Nobody opens a budgeting app because they’re excited about its AI integration. They want to know where their money went.
These are simple needs. But instead of getting better at meeting them, we keep adding more features, more configuration, and more complexity. Somewhere along the way, we confuse improving the software with increasing the amount of software.
Those sushi masters remind me that mastery doesn’t mean addition. Sometimes it means spending years getting better at using what we already have.
Maybe we software engineers should take the hint.
Get the data.
Present it clearly.
Make it easy to change.
Then get out of the way.
More
Parts Unknown, Tokyo (hbomax.com)
Rich Hickey Clojure documentary (clojure.org)
My calendar app (compasscalendar.com)