The Graveyard

How many ideas die in the space between waking and coffee?

I don’t mean that in the productivity-guru sense—capture your morning thoughts, keep a bedside notebook, optimize your ideation pipeline. I mean something darker. How much potential simply evaporates because the distance between having an idea and making it is so vast that most people never cross?

The Tax Nobody Talks About

I’ve seen it happen repeatedly. Someone with a legitimate product concept spends months learning React just to get a landing page up. By the time they’re competent enough to build, the enthusiasm is gone. The idea gets filed somewhere between “maybe later” and “who has the time.”

Or a business analyst sketches an internal tool that would save their team real hours every week. But they’re not a developer. IT has a backlog measured in quarters, not weeks. The sketch stays in a notebook.

There’s a tax here that doesn’t show up in any business case. Not a lack of talent. Not a lack of vision. Just… distance. The gap between seeing something clearly and being able to build it.

What Happened This Week

Vercel shipped something called Agent Skills—essentially a package manager for AI coding agents. On the surface: developer tooling. Another framework. Another abstraction.

But what caught me: they packaged ten years of React and Next.js optimization rules into installable modules. Rules about memoization. Accessibility patterns. Performance tuning. The kind of knowledge that takes years to develop through trial and error.

They made engineering judgment installable.

Not code. Judgment.

The Shift

Here’s where my head is.

The old path: You have an idea. You learn to code, or you find someone who can, or you hire someone, or you wait. The implementation phase is where most ideas go to die—not because they were bad, but because execution requires capabilities that aren’t evenly distributed.

The emerging path: You describe what you want. An agent applies accumulated wisdom—patterns, guardrails, optimizations—and produces something functional.

The human stays in the vision layer. The machine handles the deterministic parts.

I started out thinking this was about making developers faster. I’ve been sitting with a different question lately: what happens when the people who aren’t developers can finally cross that gap?

The Uncomfortable Part

I know what some people see when they look at this: an attack on expertise. If agents can apply ten years of accumulated judgment, what happens to the people who spent ten years accumulating it?

But I keep circling back to a different frame.

What if the bottleneck was never talent? What if it was access?

There are people with genuine product instincts who will never ship because they can’t clear the technical bar. Domain experts who know exactly what software should exist in their field but can’t build it. Ideas in notebooks, on whiteboards, in voice memos—fully formed visions with no path to reality.

What Actually Changes

If this plays out, we might be looking at something like what happened with literacy. For most of history, if you had an idea worth sharing, you needed access to scribes, then printing presses, then publishers. Each layer that fell expanded who could participate.

We’re not at another transition for sharing ideas—the internet handled that. This might be a transition for building them. Turning vision into artifact without the years of technical apprenticeship that used to be required.

The person with the idea becomes the driver again.

What I Haven’t Figured Out

I don’t want to oversell this. Agent Skills, today, is a developer tool. It mostly helps developers work faster. The “anyone can build” future is still more promise than reality.

And there’s something I’m still chewing on: when expertise becomes invisible infrastructure, do we lose the understanding of why it matters? When optimization rules get applied automatically, does the craft dimension of software erode?

Maybe. I don’t know yet.

But that graveyard keeps pulling me back. All those ideas that died not because they were bad, but because the execution gap was too wide. If we can narrow that—even partially—we unlock something.

Not replacement. Expansion.

I keep wondering what ideas are sitting in your notebook right now. The ones that feel real but impossible. What would change if the distance collapsed?