The Universal Itch: How AI is Completing the Work Shareware Started

From Shareware to “Vibe Coding,” the history of software is the history of who gets to solve their own problems. For the first time, the answer might be “everyone.”

In a previous post, I argued that Open Source Software (OSS) is not just a license, but a “culture-method”—specifically, the Itch-to-Scratch (ITS) method. As popularized by Eric Raymond, the core tenet is simple:

“Every good work of software starts by scratching a developer’s personal itch.”

For decades, this model has built our digital infrastructure. But it has always carried a fatal flaw, a structural limitation I have written about before: The Scratch Gap.

If you have the itch, but not the coding skill, you cannot scratch. You are relegated to the role of “User”—a passive entity waiting for a technical priest to feel your pain.

But when we look at the timeline of software production—from the shareware of the 1980s to the AI agents of 2025—we see this gap closing. We are witnessing the reinvention of the Itch model, where the definition of “builder” is finally expanding to include the people who can’t code—the vast majority of the population we previously dismissed as just “users.”

Shareware: The Proto-Itch

Before the Open Source Definition existed, there was Shareware.

In the 1980s, the “itch” was personal and distribution was physical. A single person (like Jim Button or Phil Katz) would encounter a data storage problem or a compression issue. They wrote the fix. They put it on a floppy disk or a BBS. They asked for $10 if you liked it.

We often look at Shareware as a business model, but it was really a production methodology. It was the first successful deployment of the Itch model at scale. It proved that valuable software didn’t need to come from a corporation; it could come from a single person solving a single problem.

However, Shareware was lonely. It lacked the collaborative tissue that would later define Open Source. If Jim Button stopped coding, the software died.

Open Source: The Collective Itch (and the Developer Wall)

Open Source took the Shareware impulse and networked it. It turned the “Itch” into a collective pursuit.

As I noted in Is OSS a SDLC Methodology?, this excelled at “Code-as-Infrastructure.” Developers are great at solving developer problems (databases, web servers, compilers).

Is OSS a SDLC Methodology?

But Open Source solidified the caste system of software. It created a culture where technical meritocracy reigned supreme. If a doctor had an idea for better patient management software, they couldn’t build it. They couldn’t “scratch.” They had to petition a developer to do it for them.

The ITS model became: Developer Itch → Developer Fix.

The “User Itch” was left largely unaddressed by the Open Source community, creating a vacuum filled by proprietary SaaS.

Micro-SaaS: The Sustainable Itch

In the 2010s, we saw the rise of “Micro-SaaS” (pioneered by figures like Rob Walling and Tyler Tringas).

This was arguably a return to the Shareware ethos, updated for the API economy. One person, one problem, one recurring revenue model.

Micro-SaaS is simply the Itch model applied to niche B2B problems. A freelancer builds a billing tool because existing ones suck. A designer builds a plugin because Photoshop is too slow.

But structurally, Micro-SaaS was still bound by the laws of the Open Source era: You still had to be a coder. The “Indie Hackers” were, by definition, hackers. The domain experts—the accountants, the writers, the teachers—were still locked out of the production room.

AI and “Vibe Coding”: The Universal Itch

This brings us to 2025. The arrival of generative AI and what the community is now calling “Vibe Coding” represents the first structural break in the Itch model since 1999.

For the first time, the “Means to Fix” is being decoupled from “Knowledge of Syntax.”

We are seeing the rise of what I call Expert Driven Development (EDD).

• A publishing product manager who knows nothing about React can build a layout engine.

• A logistics manager who knows nothing about SQL can build a routing database.

The AI acts as the technical co-founder. The human provides the Itch and the Context; the AI provides the Syntax.

This is not just “no-code” (which was often rigid and required learning complex visual abstractions). This is the ability to translate intent directly into software.

The Collapse of the User/Builder Binary

The “Itch to Scratch” model is the oldest idea in software.

• Shareware proved it could be profitable.

• Open Source proved it could be collaborative.

• Micro-SaaS proved it could be sustainable for individuals.

• AI is proving it can be universal.

We are moving toward a world where the rigid binary of “User” vs “Developer” dissolves. If you have a problem, and you have an AI agent, you are a Builder.

The barrier to entry is no longer learning to code. The barrier is now simply having an itch worth scratching.