> ## Content Index
> Fetch the complete content index at: https://madewithlove.com/blog/llms.txt
> Use this file to discover other available public pages before exploring further.

# The harp and the rain
- URL: https://madewithlove.com/blog/the-harp-and-the-rain/
- Published: 2026-11-09T09:43:00.000Z
- Updated: 2026-11-09T09:43:00.000Z
- Description: Ambient music started with a quiet record, a dead speaker and rain. Synthesis has always meant connecting what's already there; the hard part was making it. With setup decided once and Claude shaping the pieces, the gap between seeing an idea and running it has closed.
- Author: Geoffrey Dhuyvetters
- Tags: AI, Claude, AI development, Ideas, Tooling, Processes

In 1975, Brian Eno was stuck in bed after being hit by a taxi. His friend Judy Nylon put on a record of eighteenth-century harp music and left. Too quiet. One speaker dead. Rain outside. He couldn't get up to fix it.

So he listened. The harp and the rain blurred into something he'd never heard, and that afternoon became the seed of ambient music.

**Nothing in that room was new. A record, a broken speaker, the weather. The connection was.**

Eno already had a name for that kind of work. In Roxy Music he wasn't there to play the parts. He was the synthesist: the one who takes what's there and turns it into something else.

You probably have a room like that. Tabs you never closed. A screenshot of a button someone got right. Half an idea that only makes sense next to another half.

I've never been deep in [one discipline](https://madewithlove.com/blog/the-full-stack-enigma/), but I can see how pieces fit before I can say why. **Seeing was never the problem. Making was.**

An idea needs a diagram, [an interface](https://madewithlove.com/blog/no-designer-no-problem-with-v0/), some code, some copy. Some of that I do well. All of it at once? Rarely. And the ideas I did start often died in setup: picking a stack, scaffolding the project, wiring the basics. By the time I could build, the spark was gone.

The harp was playing. I couldn't get out of bed.

There's a dice game attributed to Mozart, the *Musikalisches Würfelspiel*: a waltz assembled from 176 pre-written bars. Roll two dice, the roll picks the bar, repeat sixteen times. **It always works, because the hard decisions live inside the fragments.**

My playbook for quick prototypes works the same way. **Setup, patterns, the steps from idea to something running: decided once.** The burden of starting is gone. I just pick the bars.

Then I shape. In a tool like [Claude's Artifacts](https://madewithlove.com/blog/not-rebuilding-a-slide-deck-with-pencil-dev-and-claude-design/), I bring the pieces, talk through what I see, and something appears on screen. Too busy? Strip it. Wrong feel? Swap a reference. It doesn't feel like handing work off. It feels like crossing from one craft to the next without stopping at every border.

Borrowing is old, of course. Designers have always kept inspiration folders, and Japanese garden builders named the idea centuries ago: shakkei, borrowed scenery. You don't own the mountain. You trim the hedge to frame it.

**What's different is that the folder talks back.** I open an interface in Chrome, Claude and I pull it apart, and the mountain I keep ends up in something running.

The harp is still playing. These days, I get out of bed.

I'm a synthesist.