There's a question I keep coming back to when starting any project: What kind of thing am I actually making here?
Not what it looks like. Not what it does. But what category of effort is this? Because getting that wrong means either over-engineering something simple or under-building something that needs real foundations.
Two houses help me think about this.
The Masterwork
Frank Lloyd Wright's Fallingwater. Cantilevers that engineers said couldn't work. Concrete slabs stretching over a waterfall because Wright wanted them to, physics be damned. Every beam calculated, every angle considered, every detail bespoke.
The best engineers. The best contractors. Everyone's an expert, everyone's essential. The client, Edgar Kaufmann, was told it would cost $35,000. It cost $155,000. In 1937 dollars.
The house gets built. It's majestic. A masterpiece. It's also literally falling apart. The cantilevers started sagging almost immediately. The house has needed continuous structural intervention—millions of dollars in repairs, steel cables added decades later to keep it from collapsing into Bear Run.
Wright was a genius. He was also a control freak building something that had never been built before. That's what it takes sometimes. Novel problems require novel solutions. But most problems aren't novel.
The Catalogue
Then there's the Eames House. Case Study House #8. Designed by Charles and Ray Eames—designers, not architects.
The steel frame was ordered from a catalogue. Industrial windows, the kind you'd find in a factory. Standard components, off-the-shelf parts. The original design was completely different, but when the materials arrived, Charles looked at what he had and redesigned the whole thing to use less steel. The house was assembled in a day and a half.
It's one of the most celebrated pieces of modern architecture. People make pilgrimages to see it. It's been standing since 1949 with essentially no structural problems.
Same reverence as Fallingwater. Completely different approach.
The Difference
Wright bent the world to match his vision. The Eameses bent their vision to match what was available—and made something extraordinary within those constraints.
Both houses are masterpieces. But they represent fundamentally different philosophies about how to make things.
Fallingwater asks: What would be perfect, and how do we engineer reality to get there?
The Eames House asks: What's already here, and what can we make with it?
Most Projects Are Eames Houses
Here's what I've learned: almost everything I work on is an Eames House. Not because it's less ambitious, but because the ambition is in the selection and assembly, not in the engineering.
When I'm building a website, I'm not inventing new rendering engines. I'm picking the right framework, the right components, the right approach—and arranging them with taste. When I'm designing a brand, I'm not inventing new colors. I'm choosing which ones, in which combinations, for which reasons.
The catalogue has gotten enormous. AI has made it bigger. There are more components available, more pre-solved problems, more things you can order and assemble.
This doesn't make the work easier. It makes the selection harder. When you can do almost anything, knowing what to do becomes the entire job.
When To Build Fallingwater
Sometimes you do need to build Fallingwater. When the problem is genuinely novel. When the constraints are so specific that nothing off-the-shelf fits. When you're trying to prove something can be done at all.
I'm not building Fallingwater. I'm browsing catalogues and assembling. And maybe me or someone else will assemble something new from the parts.
That's not a consolation prize. That's the Eames House.