Your Next App is Built for an Audience of One

A client asked me to create a lengthy document with manuals for various departments. The kind of thing that takes weeks to write and lives on a dusty SharePoint folder forever. Instead, I vibe-coded a user-friendly website with an AI chatbot and OAuth in about the same time. The difference? The client's employees actually used it. They could ask questions, get answers, and never scroll through 200 pages looking for "that one procedure."

This isn't an isolated case. I've recently built a pitch competition platform (from application form to jury scoring), an inspiration brainstorm roulette tool, and an events calendar. Each born from the same frustration: clunky workflows involving spreadsheets, database tools, forms, and endless copy-pasting.

We've entered an era where building software is faster than writing about software.

N equals one

Something remarkable is happening. A husband spends a few hours building an app that helps his wife choose the perfect outfit. Not for the App Store, not for thousands of users, just for her. N equals one. Others are discovering that building your own interface takes only 20 minutes. The point isn't scale, it's precision.

I've been experimenting with this myself: building a PersonalOS using Claude Code Skills. In my case, there are 28 skills and multiple instructional files steering the system. After many iterations, it produces outputs exactly like I want them. Admittedly, it took significant configuration. As with everything, the quality of outputs depends on the quality of context you put in. I consider this PersonalOS to be continuously evolving, like version control: v1.0, v1.1, v1.2.

The message here is clear: these tools enable people, even those without a programming degree, to build software that's 99% what they need. And in the case of personal systems, basically 100% personal.

From mass-market to disposable

This trend points toward something I wrote about last year: disposable software. "Traditional software installation and navigation are giving way to instantaneous, intent-driven tools that materialise when needed and dissolve once their purpose is served."

The implications ripple across stakeholders. For individuals, this means tools perfectly fitted to your quirks, workflows, and needs. For enterprises, it means rethinking IT: instead of one-size-fits-all systems, imagine employees spinning up custom tools for specific projects. For traditional software vendors, it's an existential question: why pay for bloated features when you can build exactly what you need? And for developers, paradoxically, this creates more work, not less: someone needs to configure, maintain, and evolve these personal systems.

Where only your imagination is the limit, the nature of software itself transforms. It becomes less product and more process. Less purchased and more grown.

Now imagine this for health

Picture "Claude Health Skills" as user-configurable mini-agents, each loaded with specific expertise and instruction sets, carefully designed to interact with one another. They generate 100% personal advice to lead a healthier, more dynamic life. Configurable agents that evolve together with you, similar to versions in software development.

| Traditional Health App | Personal Health Skills | |---|---| | Built for millions | Built for you | | Generic recommendations | Contextual, evolving | | Vendor-controlled updates | You control the versions | | Data locked in silos | Data flows between your agents | | You adapt to the app | The app adapts to you |

Soon we'll give citizens 100% personalised health tools: personalised dashboards, interfaces, reports, nudges, and coaching. And by then, we'll need more health professionals, not fewer. Professionals who help citizens interpret the information these tools produce and assist them in their journey toward optimal health, including setup and configuration.

The irony is delicious: the more personal our technology becomes, the more we need human guides to help us use it.

Next experiment

Identify one repetitive health-related decision you make weekly (meal planning, workout selection, supplement timing). Spend two hours configuring a simple AI assistant to help you make that decision. Document what context it needed to become useful. That's your first step into N=1 health tech.