Vibecoding got loud quick
In what feels like a matter of months, tools like Claude, Lovable, Cursor, Replit, and others moved from niche curiosities into mainstream internet culture. Alongside them came a flood of screenshots, tutorials, startup pitches, doom predictions, overnight success stories, and an equally immediate backlash from developers, designers, and technical communities.
Some of that backlash is understandable. A lot of it is reactionary. Most of it, I think, is a response not to the tools themselves, but to the loudest people using them.
When new technology becomes suddenly accessible to ordinary people, there is a chaotic period where experimentation, hype, misunderstanding, and genuine creativity all get compressed together into the same public space. Fortunes made and lost, booms and busts, and no shortage of wackiness.
People love a good story, whether they end in spectacular success or spectacular failure. Rarely do you hear about the innovations that earned a basic living. Most technologies spend far more time helping ordinary people than creating overnight success stories.
Vibecoding feels like it is going through that phase right now. Everyone is experimenting at once, often very publicly, and the internet tends to amplify the strangest or most obnoxious examples first.
The Home Camcorder Analogy
In 1983, Sony and JVC released their first consumer camcorders onto the market. What was once a technology only available to film and television studios was now in the hands of the general public.
Suddenly, a huge number of people could create something that previously required specialized knowledge or expensive equipment. Most of what got produced early on was rough, overly ambitious, or unserious. Some people treated it as a novelty. Some thought they were about to become filmmakers overnight. But mixed into all of that noise were people quietly using the technology for things that actually mattered to them.
Vibecoding and filmmaking aren’t explicitly similar, but the arrival of both onto the broader world represents a moment when a specialized activity became accessible to ordinary people at scale. Once that happens, the culture around the technology changes almost immediately.
The arrival of the home camcorder did not instantly create filmmakers. Most people were not suddenly producing carefully structured documentaries or technically impressive films. They were recording birthday parties, experimenting with transitions, filming their pets, or making strange low-budget projects with friends because they finally could. A huge amount of it was messy, awkward, and technically rough, but that didn’t make the technology meaningless. It simply meant that accessibility arrived before standards, expectations, and cultural maturity caught up. The tools spread faster than the knowledge surrounding them. Vibecoding feels very similar right now.
A lot of what gets produced with these tools today is chaotic. Some of it barely works. Some of it is built around unrealistic startup fantasies. Some of it exists mostly for screenshots and social media engagement. But mixed into that noise are people quietly building useful things for themselves, often for the first time.
That distinction matters because accessibility and expertise are not the same thing, nor do they need to be.
Lowering the barrier to creating software does not eliminate the value of engineers any more than the home camcorder eliminates the value of cinematographers. Professional work still requires experience, discipline, architecture, maintenance, and technical depth. What changes is who gets to participate at smaller scales.
That smaller-scale participation is the part I think people are underestimating.
The Loud Minority Problem, a.k.a. The Squeaky Wheel
One of the most difficult things about evaluating a new technology is that most people never encounter the average user. Whenever a new technology becomes available, average users don’t really know how to engage with it. It takes a bit of ramp up to see what something is capable of. What is encountered are the loudest users.
This seems especially true in the age of social media where drama and sensationalism drive engagement. The college student quietly building a study tracker isn’t the one getting attention. The person announcing to the world on X that they replaced an entire engineering department over a weekend almost certainly is.
The loudest users become the de facto public face of a technology, regardless of how representative they actually are.
Vibecoding is especially susceptible to this. If you spend any amount of time on X or Reddit, you’ve probably encountered some variation of the same claims: someone announcing that software engineering is finished, someone claiming they built a startup in twenty minutes, or someone treating a prototype as though it were a production-ready system despite a backend that looks like a blanket fort, if it exists at all. The veracity of these claims is largely secondary. Bold, extreme claims are memorable and spread quickly.
This is nothing new for the internet, of course. Quieter, more pragmatic instantiations find niche corners of appreciation and remain largely invisible to the broader culture, which creates a very distorted picture of how a disseminated technology is actually used.
Most people are not trying to replace billion-dollar companies, but those stories gain less traction than the stories of users making grand declarations about creating the ultimate disruptor in less than 48 hours.
When enough people encounter only the loud minority, the minority, unfortunately, begins to look like the majority.
The mascot gets mistaken for the movement.
Startup Culture and Mistaking Access for Expertise
A lot of the conversation around vibecoding has, understandably, become tangled up with startup culture. New tech, new frontiers, a gold rush on a new mountain with brand new gear. The big investors bought the mountain and distributed maps, so it’s no wonder that the mountain is covered in mining camps. All it takes is a single success for a trail of prospectors to arrive with stories of easy gold in their heads.
Almost immediately after these AI coding assistant tools became available they were swept into the same ecosystem of overnight millionaire success and passive income schemes. Amid promises of rapid and easy wealth, the lessons of the goldrush were tossed away like so many tailings: the largest fortunes were made selling tools and supplies to prospectors, not by panning for gold themselves.
With these new tools, the unfortunate first question among too many users is “how do I monetize this?” rather than “what can I build with this?”
When the focus becomes speed and monetization, expertise can start to look less important than it actually is. This is where a kind of hostility towards professional engineers emerges.
If a user completely unconnected to engineering can put together an application in an afternoon with just a few instructions, they might find it tempting to conclude that development was never really all that difficult in the first place. They can look at their own prototype and wonder what all the fuss has been about. Suddenly, years of engineering experience can look unnecessary from the outside.
A generated application might demonstrate an idea, and might even be genuinely useful. What it does not automatically provide is architecture, security, maintainability, scalability, testing, or long-term support. Those concerns do not disappear simply because creating a prototype has become easier.
Access and expertise are not the same thing.
The home camcorder did not eliminate filmmakers. The spreadsheet did not eliminate accountants. Giving more people access to software creation does not eliminate the need for professional engineers. It simply expands participation and the range of ideas that can be turned into a working solution.
The problem is not that more people can create software. The problem is the assumption that access to a tool automatically grants mastery of a discipline.
Beyond the Backlash
Looking at the discussion around vibecoding online, it's easy to come away with the impression that the technology itself is the problem.
Most of the criticism seems directed at a relatively small set of highly visible behaviors: exaggerated claims, startup fantasies, and the mistaken belief that access to a tool is the same thing as expertise.
New technologies often go through this phase. The loudest examples become the public image, while quieter and more practical uses receive far less attention.
That leaves an interesting question: if vibecoding isn't best understood as a shortcut to startup success or a replacement for professional engineering, what is its actual potential?
Top comments (0)