A developer is more than their code

This year, AI has cemented its status as an indispensable tool for any kind of software developer. The change is so major that, to a lot of people, manually writing code already seems as outdated as, say, hieroglyphs.

However, does this mean that the efficiency of AI automatically makes any software developer who uses it a better developer?

If developers delivered value exclusively by writing code, that might be a reasonable argument to make. In reality, though, a developer is much more than the code that they produce.

The real qualities that make a good developer

Coding is just one aspect of the daily work of a developer. Real projects with real-world impact require qualities that go beyond simply producing lines of code, qualities which AI cannot easily replicate due to their being uniquely human.

As AI continues to make writing code easier, developers need to focus on honing such soft skills to grow in their careers and consistently bring value to the projects they work on. These include:

• Proactive problem solving: breaking down bigger problems into smaller units and approaching them analytically and creatively
• Clear, effective communication: uncovering the real needs of a project and translating them into technical solutions which they’re able to clearly explain and document
• Collaboration & teamwork: actively listening to and engaging with team mates by sharing their knowledge and giving feedback
• Growth, learning-oriented mindset: actively following new developments and innovations, treating challenges as opportunities for growth rather than as obstacles
• Architecture & systems thinking: seeing the bigger picture while understanding true business needs, as well as focusing more on product thinking, systems design and upholding quality standards
• Taking ownership of their work: taking accountability for the code they produce, ensuring that it’s properly tested, accessible, secure, readable and easily maintainable

Since AI is now an essential part of most development workflows, a great developer also needs to be smart and responsible about their use of AI. This includes being mindful about how they share data with AI tools, leveraging AI for research, carefully reviewing the proposed solutions, setting up proper guardrails, etc.

Most importantly, a great developer doesn’t use AI as a crutch, but as a helpful assistant that makes it easier for them to focus more on non-coding aspects of their role such as architecture, strategy and stakeholder communication.

Growing the seniors of the future

Using AI smartly rather than as a crutch is particularly important when training and mentoring junior developers. While seniors have already developed the soft skills outlined above such as problem solving and big-picture thinking, juniors just starting their careers typically don’t have the real-life experience that would support this kind of growth.

Here lies the main risk of using AI in software development, which many experts have expressed concern about: if AI is so effective at development tasks that typically fall on juniors and allow them to grow, what will happen to the future development workforce if businesses rely more and more on AI over juniors in an attempt to reduce cost and move fast?

We’re already seeing the trend of downsizing development teams and reducing hiring for entry-level development roles in favour of AI. It makes sense in the moment, but represents potentially drastic future repercussions.

At Agiledrop, we’ve always respected the importance of properly growing junior developers. For the past decade we’ve prioritised hiring juniors with little prior experience and having them go through a 1-3 month structured onboarding through which they acquire practical development skills under the guidance of dedicated mentors.

Recently, however, we saw certain trends that arose from the introduction of AI into development workflows. The juniors that relied too heavily on AI during onboarding experienced significant disadvantages:

• They learned less deeply and retained less of the knowledge they acquired
• They avoided the exact challenges that would help them develop a deeper understanding of problems and solutions
• They lost the joy of overcoming obstacles to learn and create something on their own
• They collaborated less with team mates because AI already provided adequate answers

Because of this, we realised we needed to restructure our onboarding process to help our new hires avoid overly relying on AI and actually develop the skills that will serve them in the long run.

Learn more at DrupalCon Rotterdam 2026

If you’re interested in learning more about how we redefined our onboarding for the AI age, our development manager Janez Zibelnik will be giving a talk about it at the upcoming DrupalCon Europe in Rotterdam.

The session is taking place on Tuesday, 29 September, from 17:35 to 17:55 CEST (local time) in Mees Room I. If you’re attending the DrupalCon, you’re warmly invited to attend – especially if you’re also considering redefining your onboarding process to accommodate responsible AI use.