The future of software development lies in human-centred design.
Anyone who believes that AI will not profoundly change software development is, frankly, living under a proverbial rock. The transformation is already underway, with AI rapidly reshaping how software is written, tested, and deployed. This raises a question that keeps many of us up at night: what is the role of people in this AI-driven future?
I approach this question from two perspectives: industry and higher education. I’ll share my thoughts on the higher-education perspective in a future post. For now, I want to focus on industry.
For several years, AI has been assisting developers through tools such as GitHub Copilot. These tools reveal something important about programming as it has traditionally been practiced: much of the code developers wrote was boilerplate rather than deeply creative problem-solving. Syntax, configuration, scaffolding, and repetitive commands were simply mechanisms for expressing intent to machines.
“Vibe coding” is now commonplace. People with little-to-no programming experience can build their own applications. As these tools continue to improve, we can imagine a near future where operating systems and applications dynamically assemble themselves in response to natural language instructions.
Software created this way may not always be elegant or architecturally perfect, but in many cases, it will be good enough. It will be functional, immediate, and capable of solving a specific problem in a specific moment.
So why would we need software developers at all?
The case for human-centred design
The real world is messy and filled with squishy humans. We have invented organisations, cultures, social practices, and informal processes to navigate it. People rarely follow neat workflows. Instead, they develop workarounds, shortcuts, and habits that reflect the realities of their environments.
Understanding these behaviours requires observation. It means talking to people, watching how they actually work, and recognising the small details that shape everyday interactions with technology.
For the foreseeable future, AI cannot step into the physical world and observe how we behave and operate. It can analyse code, read documentation, and synthesise vast amounts of information, but it cannot sit beside someone at their desk and notice the sticky notes on their monitor, the spreadsheet compensating for a broken system, or the subtle frustrations shaping how they interact with software. AI generalises based on what it has been fed.
This is where the opportunity lies.
In the future, the value of software professionals will increasingly come not from writing code, but from observing and understanding the human contexts in which software operates. Human designers can be part of the world they are designing for—they can observe it, participate in it, and interpret it. This requires a different skill set: observational ability, design thinking, and analytical insight.
This is where human value persists in the software development process.
The days of developers building spreadsheets, calculators, and simple databases for clients are numbered. Businesses will increasingly “vibe code” these tools themselves. The savings from this automation will be redirected toward more expensive—but more valuable—bespoke systems. These systems will be designed by skilled practitioners who can create competitive advantage through novelty and ingenuity, grounded in careful observation of the real world and shaped through a human-centred design lens.
There’s a great scene that marks a pivotal point in the film Good Will Hunting. Will Hunting is a young savant who intellectually dominates everyone around him. He has an extraordinary memory and can quote books, theories, and historical facts on almost any topic. During a therapy session, he tries the same approach with his counsellor, Sean Maguire, analysing him through referencing texts about trauma and loss to demonstrate understanding.
Sean calmly challenges him:
“Thought about what you said to me the other day. About my painting. Stayed up half the night thinking about it. Something occurred to me. And I fell into a deep, peaceful sleep, and I haven’t thought about you since… [if] I asked you about art, you’d probably give me the skinny on every art book ever written. Michelangelo, you know a lot about him. Life’s work, political aspirations, him and the pope, sexual orientations, the whole works, right? But I’ll bet you can’t tell me what it smells like in the Sistine Chapel. You’ve never actually stood there and looked up at that beautiful ceiling; seen that.”
AI is similar. It has the world’s knowledge at its fingertips, but it lacks lived experience. It sees through a glass darkly1, patterns, outputs, and abstractions, but not the moment itself. It has never experienced the frustrations of a real, specific workplace, the improvisation of real users, or the subtle realities of human behaviour.
The ability to observe, interpret, and design for the messy reality of human life—that is where the enduring value of humans in software development will remain. And this thought leaves me in a deep, peaceful sleep.
- https://en.wikipedia.org/wiki/First_Epistle_to_the_Corinthians#%22Through_a_glass,_darkly%22 ↩︎