Using an app
Out on the run, my developer brain starts wondering about that weather app. What framework did they use? Swift or React Native? How does it pull data so fast? Why is it 200 MB large? The questions keep coming. But does everyone think like this? Probably not. Most users don't care about the technical details, they care that the app is fast, intuitive, and solves their problem.
Developing an App
When I'm building an app myself, should I think like a developer or like a user? The answer is obviously both! A good developer must understand the technology and the user's needs. But where's the line, and how do you decide which technical details are worth your time? It's not always easy. Over many years as a developer, I've spent a lot of time debating technical choices. Some that rarely affect users, and others that make a real difference.
Details that rarely affect users:
Details that often affect users:
-
Security – session management, input validation, and rate limiting that reduce the risk of breaches and data leaks.
-
Accessibility – semantic markup, keyboard navigation, and clear labels for screen readers.
-
Performance – compressed images, code splitting, and caching that keep things fast even on a slow connection.
-
Stability – error handling, testing, and logging that reduce the risk of bugs.
Some technical decisions lead to real, tangible value for users. Others are mostly just different routes to the same destination. They can be worth discussing, and sometimes they do matter for how efficiently a team works but how much do they really affect the end result?
From Tech Focus to User Value with AI
How can developers free up more time and energy for users' needs? There are different approaches, but for me, AI tools have been a big enabler. When AI can quickly generate code, optimize architecture, and fix bugs, I get to focus more on defining what I want to achieve. I can spend more time understanding problems, shaping requirements, and describing desired behaviours; for clients, for users, and for the AI itself. That's a meaningful shift.
Don't get me wrong, most technical decisions do affect the development process to some degree, frameworks and naming conventions included. Solutions need to be maintainable, by humans and AI alike, at least for now. But not every decision needs a human to make it. Especially not the ones that mostly come down to personal preference.
Of course, AI makes mistakes just like human developers do. Give it the wrong instructions and it can produce garbage. But the iteration cycle tends to be faster with AI tools. What still matters is making sure the output is quality work that actually solves the right problem the right way. And I think AI tools can genuinely help developers lift their gaze from the technical weeds and engage more directly with clients and their needs.
A Service Profession
I think of development as a service profession. I'm not obligated to use a particular technology or write code in a certain style to build good solutions. My job is to create value for users and clients in a sustainable way. If the tools I use help me do that well, they're good enough for me, whether that's AI-assisted or hand-written code. In the end, what matters isn't always the syntax. It's the result.
