Entry into a Tech carreer in a world of AI

I'm about to graduate with a master's degree, and I've spent the last few months loosely applying to junior engineering roles. And if I am being completely honest, the current tech market feels significantly different from what I expected when I started chasing this career path in university. And I am unsure whether this is for the best or the worst.

AI agents and generative AI have changed everything. I do have a degree in AI myself, and generative AI was an area of interest for me, and I definitely expected this rise. And yet, I am not entirely thrilled about it. What used to be a somewhat predictable path of learning fundamentals, building interesting projects, and slowly growing into a role with a high likelihood of employment now feels accelerated and even superficial. I still stand strong for the idea of deeply understanding code before shipping it, and I find it sad that this idea is slowly slipping away. Yet, here we are in today's world. This is my straightforward take on AI and what it now takes to break in as a junior, coming from someone in the same position.

The Current Reality

The entry-level job market has shifted significantly over the past 2-3 years. Many companies now expect juniors to deliver more and faster than ever before, from day one of employment. This is largely thanks to tools such as Claude and Cursor, and agentic AI. Traditional "learn on the job" positions have become significantly rarer.

Where before I would see job descriptions saying "strong fundamentals in X preferred" or "academic understanding of X", they shifted to "experience with AI-assisted development" and "ability to work with agentic workflows". This obviously doesn't mean that juniors aren't being hired anymore. They simply expect a different skillset and mindset from new candidates now.

I'm being honest, I am not in love with this shift in the market. I definitely prefer the slower, more deliberate pace of traditional software engineering, understanding the system inside and out. But I'm also realistic, sticking around in traditional principles and ignoring the change won't help me land a job.  The unfortunate reality is that companies are optimising for productivity and cost, and AI delivers exactly that.

What Skills Actually Matter

I'm not in awe of AI agents, but I'm also aware of their powers, and I utilise them every day. I've accepted that basic competence with them is necessary to be viable in this environment. And in my experience and point of view, these are skills that matter in this shifting environment.

Fundamentals Come First

Clean code, debugging, testing, and system design still matter a lot. AI is great at generating basic code quickly, but from personal experience, it often falls apart when generating larger systems and architectures.  Being able to understand and maintain proper architecture and dependencies is crucial to keep a project alive. Furthermore, AI is very capable of generating subtle bugs in code. Being able to review and improve on produced code is crucial. For this reason, I'm putting a lot of focus on maintaining my core development skills and keeping them sharp.

Use AI Properly

I've started using Claude and ChatGPT daily now for my own projects. I've learned how to write proper prompts and generate simple features. A skill I would consider invaluable in today's day and age. I've also learned to critically evaluate these outputs. I have developed skills such as spotting hallucinations, edge cases, syntax issues, and critical bugs. This brings us to my personal rule. Think of AI as a tool, not a replacement for yourself. I still write and refactor important parts myself when quality and performance count. And from personal experience, trying to get an AI to do some obscure and complex task takes more time and effort than simply doing it yourself.

Stay Human

AI cannot take away the human aspect of development. That is, the communicative and collaborative part of development. Taking ownership of systems, ensuring stakeholders' needs, handling ambiguity, and so on. I feel like these skills make you more valuable than any AI agent can ever be.

Final Thoughts

As someone graduating into this environment, my biggest takeaway is this: the junior role has changed and shrunk. Now it's more important than ever to be a developer that a company wants. Companies want people who can ship quickly with AI support, but they still need engineers who can think critically, take ownership, and aren't merely a middleman of an AI agent.

You don't have to be excited or thrilled about new AI agents to succeed. I am not. You just need to be competent with the tools, keep fundamentals strong, and lean into the human skills that matter more than ever.

If you are in the same boat, graduating soon and feeling uneasy about the direction tech is heading, you are not alone. I still feel uneasy about it. The key is finding the right balance between adapting to the new world while staying true to your beliefs and what you think makes you a good engineer.