Vibe coding closes the gap between idea and product. But that speed can trick your brain into building things that feel good instead of things that matter.
Claude Code is wonderful but agree it also enables feature creep! 😂 I usually create a separate experimental git branch where I can try out experimental features without breaking my working app!
Oh yeah I do that too! Although I need to get stricter at it cos I’ve found that I’ll name that branch something like a specific feature I was working on and then by the time I merge back into master it evolved into 10 additional things. It’s making me realise how bringing some of the structured approach to programming to vibe coding could actually really help level it up.
Thanks Joel! It’s such a great tool, but I have also experienced how easy it is for it to become addictive and take you down a different path than you planned when building. Which isn’t always a bad thing. But did get me thinking about how to make sure I’m focussing my time, energy (and credits) better instead of getting swept away in the excitement of the moment.
I just vibe code tools for myself so in theory they are all things that matter in that moment, although there are a few unfinished ones that suggest they weren't absolutely needed 😆 I find the thrill happens with the first 80% that gets built so quickly, then it fades when you have to do the real work and make whatever you've built actually do something useful - the polishing takes so so much longer but it is the important part..
Hahah oh yeah I get that! I agree, if it’s a tool for yourself you don’t need to overthink it and going with the flow can be a fun way to play and explore ideas. I bet you learned something along the way with those ones that you didn’t finish though? Sometimes the journey is better than the destination kinda thing.
Hi! I built a lot of apps, mostly for fun and for learning (or seeing the potentials) of tools like Lovable, Replit, etc. I haven't thought about something like "they were wrong or problematic", but I'm with you on that sentiment in the sense that I could've made them differently like added more or created them with less features so they would've worked worked better/in a specific way...
That’s great! Me too! I think if you’re doing it for learning or fun you don’t need to overthink it or worry about it. It’s if/when you decide hey I want this to become a real product with paying customers that it’s worth putting a bit more of a plan together.
Yes like you'd be more cautious and document each phase (better) if it's for professional use. When you're just learning, it's advantageous because you see how a specific tool works, and when the time comes that you have a grand idea of sorts, you'll know already what to do/how to tweak features to get better results.
This captures something subtle but important. The emotional signal of “progress” can easily get mistaken for actual value. The speed of vibe coding doesn’t just remove friction; it removes the natural pauses where reflection usually happens. Your guardrails feel less like restrictions and more like a way to reintroduce intentionality into an otherwise frictionless process.
I’m right there with you - it’s addictive and I’ve thought a lot about whether the time I’m spending is good or bad. 😬 Definitely the impact on my life is more top of mind than feature creep, but you’re right about that too!
Thinking about this more… I think replacing passive screen time (TV, TikTok) with Claude Code time is fine by me. And I’d rather be building because it’s fun than performing on social media for people’s approval! (Been thinking a lot about busyness and why I can’t slow down - maybe a post coming soon.) But I still need to balance it with other hobbies, family, work, etc for sure ❤️
Claude Code is wonderful but agree it also enables feature creep! 😂 I usually create a separate experimental git branch where I can try out experimental features without breaking my working app!
Oh yeah I do that too! Although I need to get stricter at it cos I’ve found that I’ll name that branch something like a specific feature I was working on and then by the time I merge back into master it evolved into 10 additional things. It’s making me realise how bringing some of the structured approach to programming to vibe coding could actually really help level it up.
A written spec can also be a helpful reminder of the actual project scope!
This is the first article I've read about the addictive nature of vibe coding. It's something I've really felt. Thanks for sharing.
Thanks Joel! It’s such a great tool, but I have also experienced how easy it is for it to become addictive and take you down a different path than you planned when building. Which isn’t always a bad thing. But did get me thinking about how to make sure I’m focussing my time, energy (and credits) better instead of getting swept away in the excitement of the moment.
Loved the insight
I just vibe code tools for myself so in theory they are all things that matter in that moment, although there are a few unfinished ones that suggest they weren't absolutely needed 😆 I find the thrill happens with the first 80% that gets built so quickly, then it fades when you have to do the real work and make whatever you've built actually do something useful - the polishing takes so so much longer but it is the important part..
Hahah oh yeah I get that! I agree, if it’s a tool for yourself you don’t need to overthink it and going with the flow can be a fun way to play and explore ideas. I bet you learned something along the way with those ones that you didn’t finish though? Sometimes the journey is better than the destination kinda thing.
Also, thanks for sharing your article!
Hi! I built a lot of apps, mostly for fun and for learning (or seeing the potentials) of tools like Lovable, Replit, etc. I haven't thought about something like "they were wrong or problematic", but I'm with you on that sentiment in the sense that I could've made them differently like added more or created them with less features so they would've worked worked better/in a specific way...
That’s great! Me too! I think if you’re doing it for learning or fun you don’t need to overthink it or worry about it. It’s if/when you decide hey I want this to become a real product with paying customers that it’s worth putting a bit more of a plan together.
Yes like you'd be more cautious and document each phase (better) if it's for professional use. When you're just learning, it's advantageous because you see how a specific tool works, and when the time comes that you have a grand idea of sorts, you'll know already what to do/how to tweak features to get better results.
This captures something subtle but important. The emotional signal of “progress” can easily get mistaken for actual value. The speed of vibe coding doesn’t just remove friction; it removes the natural pauses where reflection usually happens. Your guardrails feel less like restrictions and more like a way to reintroduce intentionality into an otherwise frictionless process.
The App Store review cycle is vibe coding's reality check.
You can ship a new feature in 10 minutes. Then wait 2-3 days for Apple to approve it 😂
Then discover users hate it. Then wait another 2-3 days to fix it.
Ship less. Think more. Apple will make sure you feel it either way.
I’m right there with you - it’s addictive and I’ve thought a lot about whether the time I’m spending is good or bad. 😬 Definitely the impact on my life is more top of mind than feature creep, but you’re right about that too!
Thinking about this more… I think replacing passive screen time (TV, TikTok) with Claude Code time is fine by me. And I’d rather be building because it’s fun than performing on social media for people’s approval! (Been thinking a lot about busyness and why I can’t slow down - maybe a post coming soon.) But I still need to balance it with other hobbies, family, work, etc for sure ❤️