The Tech Divide: Vibecoders vs. Status Quo Engineers

I'm an engineer with about 15 years of experience, and I'm tired of the denial I see between "vibecoders" and "status quo engineers". My journey started with an 80486DX and Turbo Pascal in Ukraine, where Pentiums were rare and expensive. My career path often feels like I've been time traveling:

• 2004: Learning Turbo Pascal and dabbling with Borland C.
• 2006: First job! My knowledge was outdated, so I learned MS Access on the fly and shipped a solution in two months.
• 2008: New job. My VBA skills were suddenly obsolete. "You can't build serious stuff in Visual Basic 6," I was told.
• 2012: New job. Big Data, distributed systems. How do you debug 100 machines? My .NET knowledge was trashed. C++11 was the new sexy.
• 2013: Back to the past with a new job at an equity fund, doing VBA again.
• 2018: Another new job, another trip back in time, supporting FPGA and Windows XP.

You get the idea. My career has been a constant back-and-forth through different tech eras. On one side, you have "kids" building sexy, cutting-edge stuff. On the other, "serious" engineers are building medical equipment with C and assembly code from the '90s. These two worlds can even coexist in a single company! I recently saw a hardware team genuinely excited about KDevelop's autocompletion and its ability to jump from a compiler error to the code—in 2025, amidst the "vibecoding" craze.

This experience has given me a strange feeling: there are two distinct worlds in tech, and both are in denial of the other's existence.

The Rocket Ship Camp

In one world, unicorns are pooping rainbows. Let's call it the rocket ship camp.

• Cofounders claim 10x to 100x productivity increases.
• Youtubers are already selling "vibecoding" courses.
• CEOs are firing people en masse due to AI.

These folks seem to be on a rapidly ascending rocket ship.

The Status Quo Camp

In the other world, let's call it the status quo camp, programmers are all about the "buts":

• "But it produces crap."
• "But it has security issues."
• "But it can't code in X tech."
• "But it won't be able to produce truly new things."

Usually, people in this world are simply happy to have autocomplete.

Somehow, both the rocket ship and status quo camps are hating on each other, pretending the other doesn't exist. Vibecoders produce code that often ignores security, scalability, and supportability. Status quo engineers don't even try LLM-assisted coding.

This makes me want to scream: STOP IT! There is a middle ground!

You're all right, to some extent.

• Vibecoding an MVP in days is absolutely amazing.
• But pretending that MVP is a fully-fledged product is not.
• Guiding Copilot like a lazy teenager to handle boring CRUD operations is fine.
• Trying to feed Copilot an incomplete Jira story and expecting it to figure things out is just lazy.

I don't believe LLMs will replace engineers. Instead, they'll elevate the demand for uniquely human qualities: taste, attention to detail, and critical thinking.

Yes, some low-quality consultancies will lose business on platforms like Freelancer.com, as product-oriented individuals can now "vibecode." In the long term, this will impact some incomes. But in the bigger picture, these individuals will be forced to step up their careers, learning security, scalability, and other skills that make them valuable in this new world.

Just calm down and approach this like an adult.

• Take the tools for a spin. See if they genuinely improve your productivity.
• Experiment. Build that small tool you always wanted but never had time for.
• Get ready for the next lap. AI will only get better at architecture, security, and scalability.
• And for the hardcore vibecoders: What if your product takes off? Do you want to be the next Bloomberg headline because your vibecoded app leaked PII data?