AI is becoming increasingly important in all aspects of our lives, and software development is being significantly affected by it. I've asked ChatGPT to interview me and write a little article, here is the outcome. 😊
When people talk about artificial intelligence in software development, the conversation often veers toward extremes: visions of machines replacing programmers entirely, or dystopian fears of code written without human oversight. But for many working developers, the reality is more subtle, more practical — and far more interesting. For one experienced engineer, AI has become less a threat and more a companion. “It’s productivity,” he says, matter-of-factly. “A very powerful tool that is changing how we produce software, and will change it even more in the next months and years. I don’t think it is going to replace software developers in the near future, though.”
AI is already a fixture in his daily routine, woven into multiple stages of the software lifecycle. Code completion is the most obvious win. His IDE’s assistant suggests code snippets as he works, sometimes hitting the mark, other times needing correction, and occasionally being ignored altogether. “Sometimes it needs more guidance — like comments. Other times it’s just easier to do it myself.”
Debugging, on the other hand, is a double-edged sword. When it works, it works fast, resolving small oversights instantly. But it can just as easily lead down unhelpful paths, proposing fixes that overcomplicate the code or clash with project goals.
When it comes to exploring new tools, frameworks, or architectural approaches, AI has become more like a research accelerator. He doesn’t outsource design decisions but uses AI in parallel with his own searches, combining results, asking follow-up questions, and digging deeper into promising leads. “I also use AI to gain knowledge of what I’m reading more quickly,” he explains. Only once he’s satisfied does he return to his familiar process of diagrams and documentation.
There are limits, though. Testing is one. While AI generates decent unit tests, his preference is for complex integration testing across multiple components — a task that still demands deep understanding of requirements and system behavior. For now, at least, that remains a human job.
How does he sum up AI’s role? As a brainstorming partner. “It gives a lot of relevant suggestions, but I need to put everything together myself in a way that aligns with the project vision and technical and non-technical requirements.”
Perhaps his sharpest insight is how AI pulls performance toward the middle. “If you’re below average in a domain, AI will make you average. But the opposite is also true: if you are above average in a domain then AI can bring down your work to be average, if you don’t intervene.”
That paradox is both empowering and dangerous. Juniors can climb the learning curve faster than ever before, delivering value earlier in their careers. Seniors can offload repetitive tasks, focusing on higher-level design and architecture. But overreliance risks flattening output to the median, eroding the expertise that makes great software possible.
This is why judgment and responsibility remain non-negotiable. “AI should be treated as a very knowledgeable but dumb colleague,” he insists. It can accelerate, inspire, and automate, but it cannot care. It cannot weigh trade-offs, understand business needs, or ensure long-term maintainability.
He believes developers must use the time AI saves not to coast, but to grow. The real opportunity lies in expanding knowledge, improving architectural skills, and tackling specialized domains where human insight remains irreplaceable.
“Both juniors and seniors should aim to improve their knowledge, elevate their point of view, be better architects, use the time saved from coding to expand their knowledge.”
It’s a vision of AI not as replacement, but as translation: turning human ideas into code more quickly, freeing space for deeper thinking.
The impact of AI doesn’t stop at development. It’s already transforming how people access information. Asking a model for answers is faster than combing through blog posts or technical articles. But he wonders about the long-term consequences of this shift.
“It’s hard to imagine what it will cause and how people are reacting to it,” he says. Malicious uses of AI are already widespread, but that, in his view, is inevitable with any powerful tool. The only defense is knowledge. “The countermeasure is getting to know it and learning how to deal with it as soon as possible — and this is the responsibility of each one of us.”
Looking forward, he is skeptical of the dream of artificial general intelligence (AGI) in development. In his eyes, the most valuable path is not creating autonomous, all-purpose intelligences, but building specialized assistants that excel at specific, bounded tasks.
“I don’t think an autonomous AGI could be useful if not for its own goals,” he explains. “I think specialized and reliable AI assistants would be the best thing — never with the goal to take away judgment or responsibility from people, but instead to enable and amplify people’s capabilities.”
The future of AI in software development may not be glamorous or apocalyptic. It is, instead, pragmatic: an amplifier. Used wisely, it boosts productivity, accelerates learning, and sparks new ideas. Left unchecked, it risks mediocrity and overdependence.
For this developer, the equation is clear: AI is best understood not as a rival, nor even as a partner, but as an amplifier — one that can make average developers better, and skilled developers faster and more impactful, provided they remain in control.
I believe AI is a tool, and everyone can benefit from it, maybe not in the way it’s often marketed, but still as a genuinely useful tool. Every serious professional can, and should, learn how to include it in their toolbox and make an effort to stay up to date in this rapidly evolving field.