I Shipped ListingVid While Everyone Was Talking About Vibe Coding — Here's What Got Lost in the Noise
Vibe coding is having a moment. YC is all-in. Startups are raising billions. Your feed is full of "I built a SaaS in 48 hours" threads.
Here's my contrarian take: the hype is about the wrong thing.
What's Actually Happening
Vibe coding — describing what you want in plain language and letting AI write the code — is real, and it works. Tools like Cursor, Bolt, and Lovable have genuinely compressed the gap between "idea" and "working thing." For non-technical founders or developers who want to move fast, the productivity lift is real.
The narrative, though, has gotten sloppy. "Anyone can build a startup now" conflates two very different things: shipping code and building a product that earns revenue.
What I Actually Shipped This Week
I'm building ListingVid — an AI video generator for real estate agents. This week I shipped a new feature: agents can now auto-generate a short property walkthrough video from just a listing URL and a few photos. Something that would've taken 3 days of careful dev work two years ago took about 4 hours.
So yes, the speed is absolutely real. I'm not dismissing that.
But the 4 hours of coding was preceded by 3 days of thinking. Talking to agents. Understanding why they hate editing video (it's not the time — it's the skill gap and the embarrassment of bad output). Figuring out the right default settings so the output doesn't look like a PowerPoint from 2009. Getting the CTA placement right.
AI didn't touch any of that.
Where the Hype Diverges from Reality
The "vibe coding unlocks everything" narrative assumes the bottleneck was always technical. For some people, maybe. But for most indie hackers I know, the bottleneck is:
- Knowing which problem to solve
- Finding people willing to pay for the solution
- Understanding why they're churning
- Deciding what to build next
None of that gets faster with AI. If anything, it gets more intense — because you're shipping faster, the feedback loop tightens, and the decisions multiply.
What I've Actually Learned Building This Week
- AI is a multiplier, not a shortcut. It multiplies your existing clarity. More clarity = faster shipping. Zero clarity = faster wrong turns.
- The judgment layer is where the work lives. Which feature? Which customer segment? Which pricing model? AI can't tell you. Your users barely can.
- Speed exposes your weak spots. When you can ship in hours instead of days, you run out of technical excuses. What's left is positioning, market fit, distribution. The hard stuff.
- Vibe coding is great for MVPs, dicey for production. I use it to prototype fast, then I review everything before it touches customers. Trust but verify.
- Your 15-year reps aren't wasted. Pattern recognition from years of shipping doesn't get obsolete — it gets amplified. You know faster when something smells wrong.
The Real Question
Vibe coding isn't the unlock. Clarity is the unlock. AI just makes it cheaper to test whether your clarity was right.
If you're using AI tools to go faster toward a validated idea, you're in a great spot. If you're using them to avoid doing the hard thinking about what customers actually need — you're just burning compute.
What's your experience? Are vibe coding tools helping you ship the right things faster, or just shipping faster in general?
Building in public at @lmoncany — working on ListingVid for real estate agents and OhMyLead for indie hackers who need leads without the chaos.
Top comments (0)