If I were a junior engineer entering the workforce now, I’d be a bit confused about where my career is going. There’s a conversation happening around what foundational skills junior engineers need to learn when they have AI coding tools at their disposal. There’s also a question over whether an over-reliance on AI will lead to juniors losing touch with the fundamentals of engineering.
Then there’s the party that thinks entry level engineering jobs may vanish altogether in a few years; recent data from salary platform Ravio shows roles for entry-level engineers dropped 72.2% over the last year at European tech companies — compared to a 7.4% decrease in hiring rates across all job levels.
This all begs the question: What core competencies should junior engineers develop to effectively collaborate (key word) with AI tools and remain nimble in a rapidly changing field?
I polled a few CTOs to hear their thoughts.
Emanuelis Norbutas, CTO at GenAI platform nexos.ai, tells me that the fundamental skills needed to be a good engineer are still the same, but that AI literacy is essential. “A few years ago, being proficient in one or two programming languages and understanding basic systems architecture could carry an engineer for several years. Today, that baseline also includes AI literacy — understanding how models work, how to design effective prompts, how to integrate LLMs responsibly and where their limits lie. These are no longer niche skills; they’re part of being a modern engineer.”
Understanding core computer science concepts like algorithms, data structures, system design, version control and debugging is still essential for engineers, says Marijus Briedis, CTO at cybersecurity service NordVPN — but learning how to collaborate with AI tools is equally important. Now that code can be automatically generated, it’s important that engineers can assess it for accuracy, performance and security. “Without foundational understanding, coders might miss some hallucinated code blocks or other kinds of errors,” he says.
Georgie Smallwood, chief product and technology officer at online greeting card retailer Moonpig agrees: “There will probably be a quality assurance role back in the concept of engineering soon because people will need to know how to QA at scale. The reality of AI generated code is you don’t actually know where the code is coming from. You need to be really careful about security, but there is also a question of ownership and intellectual property.”
For example, if I were to publish an article I drafted using ChatGPT that contained material collected from the web, I’m at risk of plagiarism. The outcome could be the same for engineers using AI-generated code (that may include code someone else wrote) to launch a product.
The consensus among CTOs seems clear: juniors should learn how to work with AI without expecting it to do everything for them — and that developing critical thinking and an ability to problem solve is still essential for the job of an engineer.
“While AI can assist with coding, it can’t fully generate new solutions without a knowledgeable human in the loop. A balanced approach is essential,” says Norbutas. “Juniors should view AI as an assistant, not a replacement. They should use it to generate ideas, debug and prototype, but always analyse the output critically. Developing the habit of questioning, validating and refining what AI produces is not just beneficial, it’s essential.”
Senior engineers can help their juniors by putting principles in place about how to use AI that their direct reports can follow, says Smallwood. “I want my junior engineers to attempt problems unaided first. I don’t want them to immediately code using an IDE (integrated developer environment) otherwise they won’t build those mental models that great engineers have.”
Smallwood’s advice for budding engineers wanting to make a successful career is as follows:
“Build and deploy small projects frequently to experience the full cycle of shipping code to a live environment,” she says. “Especially for those entering engineering, it's important to understand that development isn't just about writing code in tools like Lovable. Push your projects live and observe what breaks.”
Her second piece of advice is to learn how to “write clear documentation and explain complex ideas simply.” Engineers have to bridge the gap between AI outputs and human understanding, so communication skills are essential. “Cultivate the communication side of things — it doesn’t have to be verbal. The better you get at doing that, the better your AI-generated code is going to be,” she says.
Briedis advises engineers to “build durable skills, even as the tools evolve” — which they inevitably will.
“Use AI to explore and discover, but focus on writing clean and maintainable code. Learn how to debug and test deeply. Learn how to communicate clearly with both humans and machines. The best engineers I’ve seen aren’t just fast coders, they’re the ones asking: ‘Why does this work?’ and ‘What happens if it breaks?.”
