I started learning programming about six months ago after years of being an advanced Excel user. I automated much of my work with macros and later used AI-generated scripts for increasingly complicated problems, but I eventually realized I wanted to understand the code rather than blindly rely on it. I decided to focus on backend engineering because I enjoy the logic and systems side more than frontend development.
The learning process has been difficult, but I genuinely love it. Struggling with a problem and finally understanding the solution gives me an incredible sense of accomplishment. My concern is whether pursuing programming professionally is still realistic, especially as AI tools become more capable and the entry-level market seems increasingly competitive.
I'm in my mid-30s, self-taught, and don't have decades of professional programming experience or a computer science degree. I've been a musician for more than 20 years, so I understand the difference between enjoying something and being able to make a career from it. I'm happy to keep programming as a hobby if that's the sensible choice, but I'd like an honest assessment: are there still viable paths into software development for someone with my background?
4 Answers
Yes, there are still programming jobs, but the entry-level market is much tougher than it used to be. A degree can help you get past screening, but it isn’t the only route. Without one, you need something concrete to compensate: strong projects, useful experience, open-source contributions, or knowledge of a particular business domain.
Your Excel and automation background could be valuable. Look at backend automation, internal tools, data processing, integrations, reporting systems, and workflow software. You already understand how businesses use technology to solve practical problems, which is more useful than simply completing tutorial exercises.
A formal degree would improve your odds with many employers, but self-taught developers can still get hired. The process may take longer, and you’ll need to demonstrate practical ability more clearly than a graduate with internships or a recognized credential.
Contributing to open-source projects, submitting useful pull requests, building tools that other people can use, and gaining experience through work-related automation can all help. Focus less on trying to prove that AI will or won’t replace developers and more on becoming someone who can reliably deliver and maintain software. There is still demand for that, even though the job will continue changing.
AI hasn’t made software engineering pointless, but it has raised the bar. Tools can generate code quickly, yet someone still needs to understand requirements, architecture, security, debugging, testing, deployment, and whether the generated code is actually correct. Developers who can use AI effectively while understanding the underlying systems may be more productive, but that also means companies may hire fewer inexperienced people.
So don’t learn only how to prompt an assistant. Build the fundamentals and use AI as a tool you can audit rather than as a substitute for understanding.
Treat it as both a hobby and a serious career experiment for now. Keep studying, build a few complete projects, and apply what you learn to real problems instead of waiting until you feel fully ready. You could recreate some of your Excel automations as proper backend services, add a database and an API, deploy them, document them, and explain the decisions you made.
After six months, it’s too early to conclude that the path is closed. Give yourself a defined period of consistent work—perhaps a year or two—and evaluate your progress, finances, local job market, and the kinds of roles you can realistically target. You don’t need a guarantee before starting; you need evidence that your skills and opportunities are improving.

I’ll keep going regardless because I genuinely need the challenge and mental stimulation. Even if it never becomes a job, I know this is one of the hobbies that will stick. My main uncertainty is whether I should still seriously aim for a programming career as a self-taught person in my mid-30s.