The Day I Stopped Playing Every Instrument

For most of my career as a developer, building software meant playing every instrument in the orchestra.

If I wanted a new feature:

• I designed the database schema.
• I wrote the controller.
• I added the validation rules.
• I wired up the frontend.
• I created the tests.
• I wrote the documentation.

If something sounded wrong, I had to find the instrument and fix the notes myself. This was just how software development worked.

You didn’t just write the music. You also played every instrument.

Laravel Already Changed the First Part

Laravel already shifted this idea years ago.

Instead of forcing developers to start from nothing, Laravel handed us instruments that were already tuned:

• Authentication scaffolding
• Migrations
• Queues
• Starter kits
• Conventions that removed dozens of decisions

You didn’t start by building the entire orchestra. You started closer to the music (some of the instruments were already tuned by Symfony behind the scenes).

Laravel’s philosophy was: don’t waste time building the stage; start performing the piece.

AI takes this a step further.

The Orchestra Is Already There

With tools like ChatGPT and Claude Code, a lot of the setup work is simply done — fast.

I describe the feature I want. Something like:

“Users should be able to upload a profile photo. Store it, validate the file type and size, and expose it through the API.”

And suddenly there is code. A migration. A controller. Validation rules. Sometimes tests.

It’s not always exactly right. But it’s often surprisingly close.

Instead of assembling every instrument myself, it sometimes feels like walking into a concert hall where the orchestra is already seated.

Conducting Still Requires Understanding the Music

At first, I was just amazed by how fast it could produce working code.

Watching an AI scaffold a feature felt like seeing my first automated UI tests run: the characters flashing by, tests passing, migrations created, controllers and routes appearing almost instantly. There’s a small thrill in watching something that used to take hours happen in seconds.

But once the novelty of speed fades, the weight of the role shifts. You aren't just a faster typist; you're a supervisor.

I spend less time writing every line. More time:

• Shaping architecture
• Reviewing AI output
• Adjusting design
• Refining results

The code still matters. Understanding it still matters.

AI can play the instruments. But someone still has to make sure the music makes sense. A conductor doesn’t play the violin during a performance. But they know when the strings are off, the brass is too loud, or the tempo drifts.

That is the evolution: we are moving from performing every note to guiding the entire orchestra.

The Cost of Software Is Changing

Software creation is becoming dramatically cheaper.

Ideas that once took days or weeks can become prototypes in an afternoon.
Small tools that weren’t worth the effort are suddenly easy to build.

More software will exist than ever before.

But this raises a question: if software can be created so quickly, do we start prioritizing quantity over quality?

Many of these AI-generated apps are “tried and tested” only in the sense that they run once or twice. They haven’t been used in the real world, debugged under heavy use, or maintained over time. The software works, but it may be fragile, prone to breaking, and lacking the resilience that comes from careful, iterative development.

The Developer’s Role Is Evolving

AI hasn’t removed the need for developers. It has changed what we do.

Less time typing every line. More time:

• Guiding systems
• Spotting subtle problems
• Shaping software that lasts

The orchestra is bigger now. And increasingly, the developer is the conductor.

In This Series

In the coming articles I’ll explore what this shift really means:

• Vibe Coding Is Like Conducting Without Experience
• From Scaffolding to Symphonies
• Where AI Is Amazing
• The New Economics of Building Software
• Are We Entering the Era of Fragile Software?

The orchestra now includes AI. Let’s explore what that really means.