If you're just starting with coding, using AI to generate your code might sound like a shortcut to success. But actually, it could hold you back in ways you might not realize. Here’s the thing: as a new programmer, your main focus should be on learning the basics and building a strong foundation. Coding isn’t magic, and it’s not about writing beautiful poetry either. It’s about giving clear, step-by-step instructions to a computer to get real things done.

Take JavaScript, for instance. If you started learning with it, you might not have seen the lower-level side of how code is run by the computer. And that’s okay! But a lot is going on under the hood that can give you a much deeper understanding. I started with C, which grounded me in low-level programming concepts. In university, I even took a course in Assembly language. Yep, Assembly—the dinosaur language that people still use in hardware programming today, but it taught me so much about how the program works.

Adding two numbers in Assembly looks like this:

.model small
.stack 100h

.data
    num1 dw 10       ; Define a word (16-bit) with value 10
    num2 dw 20       ; Define a word (16-bit) with value 20

.code
main PROC
    mov ax, num1     ; Load the value of num1 into AX
    mov bx, num2     ; Load the value of num2 into BX
    add ax, bx       ; Add the values in AX and BX

    ; Exit the program
    mov ah, 4Ch      ; DOS interrupt for program termination
    int 21h          ; Call DOS interrupt to exit
main ENDP
END main

But in JavaScript, it’s just:

let sum = 5   3;

Or in Python:

sum = 5   3

Today, we’ve got tools like ChatGPT, Gemini, Claude, Cursor and Bolt that can churn out lines of code for you in seconds. It’s cool, but is that what you want? If all you’re doing is writing prompts and waiting for code that’s not even yours, you’re missing out. The real joy comes from figuring things out yourself—puzzling through the problem, working out the solution, and building something you own. Trust me, that’s way more satisfying.

Now, don’t get me wrong—AI tools can be great. They're super helpful for automating repetitive tasks, writing some CI/CD scripts, explaining confusing sections of code (just be careful with sensitive data!), or even brainstorming project ideas. But at the end of the day, your job as a developer is to solve problems. Crafting solutions to real-world challenges or helping to build your company’s next big product is what will make you a better developer—not learning how to write the perfect prompt.

Plus, AI-generated code still needs a human touch. If you focus on growing your skills and learning from people around you, you'll find that your growth is more meaningful and lasting. Coding is just one part of software engineering. Debugging, analyzing problems, quality assurance (QA), UI/UX design—there's a lot to this field! Relying only on AI early on means missing out on building these other essential skills, and that could end up holding you back.

Another big area to focus on? Data structures and algorithms. Without a solid understanding here, how will you know if AI's solution is efficient? Are you going to keep prompting it until you find a better answer? That sounds exhausting—and it's not the best use of your time. Pick up the keyboard, grab a coffee, and dive into the code yourself. There's nothing quite like the satisfaction of seeing your solution come to life.

At the end of the day, AI can be a helpful sidekick, but don't let it become a crutch. If you want to be a great software engineer, it'll take time, patience, and a lot of hands-on practice. AI is just a tool. The real magic comes from you.