One developer can take on more of the work.

I’ve been building software forever. AI has made it interesting all over again. A large part of that excitement is being able to try more of an idea myself.

A small software product involves several kinds of work: product management, interface design, development, quality assurance, technical writing, marketing, and customer support. I’m interested in how those jobs might change when one person can explore more of them with AI.

Ask for help with a particular job

I can ask AI to propose a clearer flow, implement it, review the change, draft instructions, and help explain who the product is for. Each request needs context and a result I can inspect.

For a design, I want to see someone complete the task. For QA, I want failures and recovery checked. For marketing, I want an accurate explanation of an implemented benefit. “Help me launch” is too vague to judge.

The handoffs still matter

A design decision needs to survive implementation. A defect found in review needs to be corrected. A product description needs to match the software. Keeping those pieces aligned is part of my work, even when AI helps produce them.

I’d rather keep a short record of decisions and a clear source of truth than hope every new conversation remembers what the last one meant.

More reach is the exciting part

I don’t need a separate person for every role in every small project. AI gives me a way to do more of that work myself. There will still be situations where specialist help earns its place.

That’s what I mean by a force multiplier: more useful work within reach of one developer. My experience helps me direct it. My curiosity keeps me trying. I want to see how far that combination can go.

— Vince