4am, 47th prompt sent. If you do Vibe Coding seriously, you know the scene. Why it's so addictive, and three reflexes to make it last.
It's 4am. My coffee is cold. I've just sent my 47th prompt. At 8:30 I have a client meeting. Guess what I do? I send a 48th prompt.
If you've done Vibe Coding seriously, you know this scene. If you haven't lived it yet, get the coffee ready: it's coming. At Dotika we've trained hundreds of people in Vibe Coding. Every cohort, I repeat the same line: 'Watch out for the dopamine.' Every cohort, someone laughs. And three weeks later, someone messages me at 2am: 'Mišo, I think I have a problem.'
Not a real problem, don't worry. The most beautiful problem in the world: building becomes so much fun that you can't stop.
Why it's so addictive (in the good sense)
You build an idea in 10 minutes. You test it. It works. You add a feature. That works too. One more. One more. It's 1am. That's normal. It's even the sign you've found something that makes you tick. Vibe Coding triggers a genuine joy loop, because for the first time in history there is no friction left between having an idea and watching it come alive.
Before, between 'I have an app idea' and 'it's running on my screen', there were six months and 5,000 lines of code. Today there are 15 minutes and three prompts. That compression of time is magical. It's also why we struggle so much to close the laptop.
And I'm not the only one living it
Vibe Coding is more addictive than any video game ever created.
The stars of the movement talk about it openly: Armin Ronacher, creator of Flask, says 'we barely sleep, we build incredible things'. Garry Tan, CEO of Y Combinator, sleeps 4 hours a night — and reckons a third of the CEOs he knows do the same. Does that reassure you? It does me. If you've chained three late nights this week, you're neither weird nor alone. The question isn't how to brake that energy. It's how to channel it so it lasts.
My shameful little story
One Sunday evening, around 9pm, I tell myself: 'I'll just quickly test something on Lovable.' At 2am I had built three apps. None useful. None finished. Monday morning I had a client board meeting at 9. I showed up puffy-eyed, unable to string two strategic sentences together. The funny part? During the meeting, I was thinking about the fourth app I'd start that very evening.
That's the trap. It's not that Vibe Coding is dangerous. It's that it's so much fun that without a minimum of discipline, you build ten half-finished apps instead of one that actually changes something.
How to know if you're already there
- You fall asleep drafting prompts in your head.
- Your collection of unfinished apps grows faster than your bank account.
- The joy is in starting a new project, not in finishing it.
If you ticked all three, welcome to the club. I'm the president.
What works for me
- Close the laptop at 10pm. Whatever comes out of your brain after midnight, you'll spend twice as long fixing the next day. 3am code is rarely the best code of your life.
- One app at a time. Before starting a new one, finish the current one or bury it properly. Tonight's genius idea will keep until tomorrow. If it doesn't, it wasn't that genius.
- Define 'done' before you start. One clear criterion, written down, to know when to stop. Otherwise you'll prompt forever, because 'a little better' always exists.
Vibe Coding is great. Really
We spend our days training people in Vibe Coding. We believe in it completely. It's a revolution, not a fad. For the first time in history, someone who has never written a line of code can build a tool that solves a real problem. But like any revolution, it requires learning to dose. Vibe Coding is a bit like excellent espresso. One in the morning: great. Eight in a day: you don't sleep.
So there you go — don't say we didn't warn you. Vibe Coding will get you hooked. That's the plan. But now you know. Close the laptop at 10pm. Finish before you start. Define 'done'. And enjoy it. That's what it's for.
Mišo Pecirep is Co-Founder and COO of Dotika. He has gone to bed at 5am more than once. He writes this article on a Friday morning, after a good night's sleep.