"Vibe coding" was coined by computer scientist Andrej Karpathy (OpenAI co-founder and former AI lead at Tesla), who described working with AI as more conversational than traditional coding: “It’s not really coding—I just see things, say things, run things, and copy-paste things, and it mostly works.” As overused as the word vibe can be, it fits here. 

Before the term existed, I was vibe coding during my undergrad thesis in Unity. I’d used Unity and C# before, but this project stretched my skills. It was Spring 2023—the first time I used ChatGPT to help code. At first, it felt like cheating, but really it was just faster, smarter coding. I still used Stack Overflow, Google, YouTube, and docs, but ChatGPT took care of the grunt work. 

I still use ChatGPT and Copilot today, and coding has never felt smoother. The quality of code and ideas these tools provide has noticeably improved since 2023. 

The main draw of vibe coding is how it lowers the barrier to entry and speeds up workflows. Beginners can now start coding and prototyping games much faster, and AI can even help generate creative ideas. Designers without deep programming skills can finally bring their visions to life, and solo devs can build larger, more polished games with fewer resources. 

Of course, vibe coding isn’t perfect. If you don’t understand what your code is doing, you’ll run into bugs you can’t debug. AI lacks full context, and its suggestions can create inconsistencies—especially in bigger projects. Beginners might become too reliant on AI, slowing their long-term growth. And creatively, there’s a risk of derivative ideas if you're not pushing past what AI already knows. 

Still, when used thoughtfully, vibe coding is a net positive. It’s made me more productive, more confident, and able to do things I couldn’t before. You could say I vibe with vibe coding.