I Revived A 4-Year-Old Abandoned React Quiz App

Temisan — Sun, 24 May 2026 14:18:36 +0000

This is a submission for the GitHub Finish-Up-A-Thon Challenge

What I Built

I revived a 4-year-old React quiz app I abandoned on GitHub.

It had been sitting untouched for years... broken, incomplete, and barely functional. When I fired it up, it didn't even make it past the loading screen.

Sigh... I expected this anyway.

I opened my browser console and, voila, I found a couple of errors:

429 Too Many Requests from the OpenTDB API
Cannot read properties of undefined (reading 'map') — the app was calling .map() on a failed API response

The API kept firing on every render with no error handling anywhere, while the spinner just ran forever. A classic codebase built by a wannabe self-taught junior developer with limited understanding of JavaScript fundamentals.

Demo

Live demo: https://tryquizzify.vercel.app/

Repo: https://github.com/temisan0x/quiz-app

The Comeback Story

And yeah... that was the starting point.

First thing I did was slow down and actually trace what was breaking.

In api.ts, I refactored the request logic:

Added a response.ok check so 429 errors don't silently pass through
Guarded data.results with Array.isArray before mapping over it
Wrapped startTrivia in a proper try/catch/finally so the loading state always resets, even when the API fails

That alone stopped the infinite loading spinner that made the app unusable.

Next up was the actual gameplay. The original version didn't really feel like a complete quiz app. It just started and ended without feedback. It was boring, to say the least.

So I added:

A difficulty selector (Easy, Medium, Hard)
A proper results screen with score feedback
A "Finish Quiz" state so users know when they're done
Small sound effects for interactions to make it feel more responsive

The old UI was functional but, to be honest, it didn't match what the idea of the app was supposed to be. So I replaced it with a darker aesthetic, a more eccentric trivia-styled theme, and glassmorphism cards to make the content stand out.

My Experience with GitHub Copilot

Copilot helped most during cleanup and refactoring.

It was useful for spotting repetitive patterns and suggesting safer ways to handle async API calls. A few times it pointed out edge cases around response handling that I hadn’t fully considered.

I didn’t just follow it blindly, but it definitely sped things up when I was tightening the data flow and fixing the loading logic.

At one point it suggested checking Array.isArray(data.results) before mapping — which ended up preventing a silent crash I had completely overlooked.

It's not a perfect app... but it's a finished one, and that's the point.

We Are Wrong About AI Coding.

Temisan — Tue, 19 May 2026 10:16:58 +0000

The past couple of years have been pretty wild for developers. Ever since AI entered the picture, the speculation has been nonstop.

"AI is taking over the entire industry. Developers are finished. The robots are coming."

Wild, right? Well, I'm still not convinced. Most of these claims are clearly exaggerated.

Let's just take the base claim that AI can write code 10x faster than any developer. Or better still, faster delivery or 10 times more code output... of which I am not convinced.

I don't think it's that simple.

Don't get me wrong, I'm not trying to play the devil's advocate here.

The CEO Thought Experiment

Take for instance, you're the CEO of a company. Your goal is to maximize profits... With the rise of AI, you can now crank out features, utilities and products like never before. All you gotta do is start using AI. Right?

So what do you do?

Option	Action	Outcome
Option A	Fire 90% of your team, use AI alone	Same output as before
Option B	Keep your team + use AI as a multiplier	Ship 10x more than competitors

Please stick with me here. This is a thought experiment. I'm not saying this is what you should do. I'm just asking you to consider the options.

Well obviously the second option makes sense. But here's the thing... if you picked Option A while your competitors picked Option B, they'd bury you. More output, more features, faster delivery. Your company would fade out and die.

So Why Aren't We Seeing 10x Improvements Everywhere?

📌 The job market for programmers is rough right now. But I don't think AI is the main reason. The entire economy is down.

Between 2021 and 2023, tech companies massively overhired during a boom. When the economy contracted, they held onto that workforce for a while, then started cutting. The AI narrative became a convenient cover story. In reality, most of these companies were just losing profit.

AI is probably getting more blame than it deserves.

Most companies firing developers will say it's to implement AI. But dig a little deeper and you'll find a bad market, shrinking margins, and cost-cutting decisions that had been coming regardless.

This Isn't the First Time

This isn't the first time a new tool has threatened to make developers obsolete either. Throughout the history of programming, new technologies have consistently made developers faster, better, more productive. And the industry grew every single time.

More capability meant more ideas, more products, more demand for developers who knew how to use the tools.

"More resources doesn't mean less work. It means bigger ambitions."
Same story. Different tool.

Where Does AI Actually Stand?

AI is a stronger multiplier than anything that came before it. I'll give it that. But I don't think the outcome will be any different.

A competent developer using AI will outbuild and outship someone who isn't. That doesn't mean less work though. It means more features, bigger products, wilder ambitions.

That's my take anyway. See you in the next one.¹

¹ Idea credit: Kyle Cook — Web Dev Simplified

DEV Community: Temisan