AI won't replace developers—it's your sidekick. Exploring the reality of AI in software engineering and how to use it effectively.
Back in 2023, artificial intelligence (AI) was heralded by some as the "developer killer". A terrifying force poised to undermine traditional software engineering - potentially rendering developers obsolete.
Fast forward to today, and this fearful narrative seems not only premature but fundamentally flawed. Instead of streamlined workflows and effortless productivity, I and many developers (and businesses) find themselves navigating a chaotic landscape where AI's promise feels at odds with the day-to-day reality.
The Chasm of Chaos
The modern software development environment is awash with AI models, each claiming to be the ultimate tool for enhancing productivity. From large language models to specialised coding assistants, the sheer volume of options can be paralysing. Add to that the limitations of small context windows, inconsistent outputs, and the struggle to integrate these tools seamlessly into existing workflows, and it's clear: the AI revolution in software engineering is far from tidy.
Developers often face what could be described as "agentic nightmares"—overwhelmed by autonomous systems that lack coherence and standardisation. This fragmentation makes it difficult to "work smarter, not harder", a promise that initially fuelled the AI hype.
A Glimmer of Hope: Moving Towards Standards
From my own personal conversations with colleagues and peers, not one of them uses AI the same way. It's a very unique experience for each person.
The good news? Some developers and industry leaders have recognised these pitfalls. Efforts are underway to establish common standards, improve interoperability, and create guidelines that help developers harness AI without the chaos. These initiatives aim to transform AI from a disjointed collection of tools into a cohesive, reliable part of the software development toolkit.
The Sidekick You Didn't Know You Needed
So, is AI the developer killer it was once touted to be? I doubt it. In its current state, AI is far from replacing human developers. However, when used correctly, it can be an invaluable sidekick—your junior developer who's always ready to assist, offering suggestions, catching errors, and automating repetitive tasks.
Based on my own experience, AI has been a tremendous buddy in peer reviews, helping me understand complex problems and providing holistic views on a codebase. While AI offers clarity and focus, especially when working asynchronously across time zones, it's not a substitute for the dynamic interaction I have with my peers. Pairing with fellow developers during peer reviews fosters a deeper understanding and maintains that vital human connection, crucial in a mostly remote role. After all, AI won't ask me what GAA jersey I'm sporting today!
Think of AI as a tool that complements your expertise rather than competes with it. It's not about replacing developers but empowering them to focus on creative problem-solving and strategic thinking while AI handles the grunt work.
Moving Forward
AI in software engineering is still evolving, but it's got some seriously exciting potential. With all the hard work and efforts to make these tools more standard and reliable, AI is becoming the super-smart assistant every developer needs. It's not going to take our jobs away—it's going to help us do them better.
The journey of AI in software engineering teaches us a valuable lesson: it's not a sign of doom, it's a chance for us to evolve and grow. Instead of fearing AI, let's think of it as a sidekick that can help us take our coding skills to the next level.
References:
1. InstituteData - 25-08-23: https://www.institutedata.com/us/blog/ai-in-software-development-can-ai-replace-software-engineers/
2. CNBC - 07-09-25: https://www.cnbc.com/2025/09/07/ai-entry-level-jobs-hiring-careers.html
3. Addy Osmani - 31-01-26: https://addyosmani.com/blog/self-improving-agents/