So I kept asking myself… are we using AI the right way?
Everywhere I look, someone is saying AI will take developer jobs. Someone else is saying it's just hype. And in the middle of all that noise, most of us are quietly pasting prompts into a chat box and hoping for the best.
So instead of guessing, I decided to learn it properly.
Over the past two weeks, I went through Google's AI Professional Certificate, from the basics of prompting all the way to building and deploying an app with AI.
Here's what actually changed how I work.
1. Your prompt is your spec.
Vague input gives vague output. Every dev knows what a bad ticket does to a project. A bad prompt does the same thing. Now I give AI a role, a task, context (stack, constraints, users) and the format I want. Then I follow up and refine instead of accepting the first answer.
2. Plan before you generate.
I used to jump straight to "build me this." Now I define what the app should do, who it's for and what "done" looks like first. Code generated from a fuzzy idea becomes a debugging nightmare.
3. AI can be confidently wrong.
It can hallucinate, use outdated methods, or write code that looks perfect and fails quietly. If I can't explain what the code does, I don't ship it. Test it. Debug it. Understand it.
4. Security still matters.
As someone deep into cybersecurity, this one hits close. Never paste API keys, secrets or client data into AI tools. Review generated code for weak validation, injection risks and exposed credentials. AI doesn't care if your app gets breached. You do.
5. Fundamentals still win.
The better you understand APIs, databases and security, the better you can direct AI, and the faster you spot when it's wrong. AI amplifies the skills you already have. It doesn't replace learning them.
I came away with one big mindset shift: AI is a collaborator, not a vending machine. You bring the judgement. It brings the speed.
For us building in Africa, that matters. With limited resources, tools like this can help us prototype faster, learn faster and solve local problems, if we use them wisely and keep growing our own skills.
Here's a small challenge for this week: pick one repetitive task in your workflow (writing tests, documenting code, cleaning data) and try it with AI. Then check the output carefully.
Come back and tell me:
- How are you using AI in your workflow right now?
- What's the biggest mistake you've seen AI make in your code?
- Does AI make learning to code easier or harder?
I'd love to hear your experience. 👇
Signature:
The Duchess of Hackers
Full-Stack Developer | Digital Marketer | Cybersecurity Enthusiast