programmingGameDevpersonal

Vibecoding Isn't Fun

RF
Ryan Fikri
4 min read

I love puzzles. Ever since I was a kid, I played a lot of puzzles. I remember that I played several Flash puzzle games such as “Missionaries and Cannibals”, “The Lonely Knight”, and “Wolf, Sheep & Cabbage” on a Toshiba Satellite A series. I forgot the model, but I think it was an A100, and it ran good old Windows XP. Then, in the 2010s, my mom bought her first Android phone: a Samsung Galaxy W running Android Gingerbread, and I played some “Angry Birds” games; then my dad bought a Samsung Galaxy Tab 2, and I played “Cut the Rope” and “Where’s My Water?”. Looking back, I think what I really enjoyed wasn't the games themselves, but the feeling of solving something.

Hello World

In 2020, during the COVID-19 pandemic, I felt bored at home. When I scrolled through Instagram Reels, there was a video of someone using an Instagram filter that randomly shows a Python snippet, and the user has to guess the output. I thought it was fun, and I immediately searched how to learn Python. Then I started taking an online course on Sololearn and coded some Python scripts, like a calculator, Rock Paper Scissors, and a simple quiz game. I remember that it was pretty fun and rewarding when the code worked as expected. Because of this, I thought about majoring in computer science at university.

Hello College

I entered college in 2022, and the first year was pretty fun. I learned about programming fundamentals, object-oriented programming, and algorithms. And I also created my first game, NonsenseNebulaProblem. It was one of my favorite projects, and it was also one of the times I had the most fun programming. I remember that I spent days to weeks scrolling through cppreference, the SFML forum (thanks eXpl0it3r for answering many questions there; I keep finding his nickname when scrolling), and StackOverflow to find how to do things, and I was very excited when I finally solved it.

Hello ChatGPT

Just after I entered college, ChatGPT was released. At first, I was skeptical about it. But then I tried it, and it was a game changer. I use it to help me with my assignments and to quickly find out how to do things. Little did I know that I’d rely more and more on LLMs as time goes by.

Hello Vibethinking

Nowadays, I started noticing that I rarely question the solutions that LLMs provide, and that made me wonder whether my critical thinking skills are declining. It’s kinda worrying, because I want to be a software engineer who can design and architect systems, not just someone who can implement things. I want to be able to understand the problem deeply, think about various solutions, evaluate them, and choose the best one. But I feel like I’m losing that ability. I rarely question the solutions that LLMs provide. It’s like I’m outsourcing my brain to them. When I come across a problem, I no longer feel any desire to try to design the solution, and using an LLM to solve it doesn’t give me the same satisfaction as before. The fun is gone.

My Two Cents

Overall, it’s a mixed bag. LLMs are powerful tools that can help us be more productive, but we need to be careful not to become too reliant on them. It’s a balancing act between leveraging their power and maintaining our own critical thinking skills. The key is to use them as a tool to augment our thinking, not to replace it. So, my message to my fellow programmers or to anyone who just starts programming: don’t skip learning fundamentals; don’t just use LLMs to generate code, but also use them to discuss the problems. Instead of asking the LLMs to solve something, try to think about it first, come up with a solution, and discuss it with the LLMs. LLMs have a wide knowledge of system design, algorithms, data structures, etc. It's a shame not to use it, but it's an even bigger shame not to use our own brains.

Btw, here's some screenshot of my old game project. I know it's crappy, but I had fun building it, so it doesn't matter :v nnp_home nnp_game