How The New “Vibe Coding” Culture May Forever Change Tech
AI-powered “vibe coding” has made it so that AI can engineer anyone’s app without human software developers. What are the ramifications?
Before we get started—there’s another Hard Reset event this week!
Are you a tech professional who wants to unplug from your screen and take a foray into nature? Join Hard Reset Media and Climate Action Club for a GUIDED NATURALIST WALK in Golden Gate Park with @thebayforager, Nick Robertson!
As the tech landscape rapidly shifts, Hard Reset & CAC wanted to bring you an event where we could chat about the overlap between tech and climate: everything from success stories to tales of caution and action plans.
We’ll begin with a two-hour guided walk to explore our local wildlife and identify edible plants. Then we’ll host lunch at 12 PM to connect as tech builders, founders, designers, and engineers who are interested in the world we can experience offline.
A few months ago, a New York Times piece was published with the title Goodbye, $165,000 Tech Jobs. Student Coders Seek Work at Chipotle. According to the article, the reason that computer science graduates are now seeking out jobs at fast food chains is because:
..tech firms are embracing A.I. coding assistants, reducing the need for some companies to hire junior software engineers. The trend is evident in downtown San Francisco, where billboard ads for A.I. tools like CodeRabbit promise to debug code faster and better than humans.
Called “vibe coding,” a term coined a year ago by AI researcher Andrej Karpathy, everyday people are now building digital products that previously may have required a team of a designer, a product manager, two engineers, and several months of design, coding and testing. The collective salaries of these people might total up to hundreds of thousands of dollars.
But today, with vibe coding, this same work could possibly be done for a few thousand dollars—having potentially seismic implications for the labor market and industry. As essayist and technologist Paul Ford wrote last month in the New York Times:
The tech industry is a global culture — an identity based on craft and skill. Software development has been a solid middle-class job for a long time. But that may be slipping away. What might the future look like if 100 million, or a billion, people can make any software they desire? Could this be a moment of unparalleled growth and opportunity as people gain access to tech industry power for themselves?
Those are thorny questions, but overall, it’s impossible to ignore that vibe-coding is now proliferating throughout not only the technical culture, but the culture at-large. A CEO of an A.I. coding startup—albeit biased—said that that the vibe-coding space is at the “bitcoin $1 moment.” Meanwhile, Chief Product Officer of fintech startup Ramp said on a podcast that employees who are not vibe coding are “underperforming,” and Amazon is promoting the voices of employees who “vibe coded” their way to a promotion.
Beyond companies building their own tools in-house to streamline coding, A.I. platforms like Replit and Lovable are making it so that anyone with an app or digital product idea can build without human software engineers. Coding platform Lovable told Business Insider this week that it’s seen at least 200,000 new vibe coding projects created each day.
Technical veterans are torn. I asked Justin Hunter, Head of Product at StableCore who has also built apps in the decentralized finance space, about the surge of vibe coding.
“There’s a lot of negativity around vibe coding among developers, especially senior developers, the assumption being that vibe coding produces slop and doesn’t actually teach programming,” Hunter told me. “But my 14-year-old son is the antidote to vibe-coding negativity. He’s been working with Claude Code for the last few weeks to build Minecraft mods and just recently earned his first money on the internet by publishing these mods online. He has dyslexia and ADHD and vibe coding has given him a creative outlet to produce more than he consumes, and he’s starting to understand languages like Java and Python.”
For Raza Khan, an entrepreneur and investor who works with a number of highly technical founders in the biotech and legal spaces, guardrails need to be put into place sooner than later. Khan told me that much of the important infrastructure meant to formally verify code is not in place for most AI-generated code, elaborating:
Vibe-coding works—until it doesn’t. In practice, today’s AI can actually make programming harder. Engineers are already overwhelmed, sifting through the relentless stream of code AI generates from a simple prompt. Sure, the output might be correct. But more often than not, it isn’t. And even when it works, maintainability becomes a nightmare. You can’t easily improve what no one truly understands. Changes that should fix things end up starting fires instead.
That said, I do see a path forward—one where people use formal methods to generate correct, safe software. It’s an area I’ve invested in, and it’s finally gaining traction. With the right infrastructure, the code I generate has to conform to my specifications. It does exactly what it’s supposed to. And it’s far easier to maintain.
Without that kind of rigor, AI-generated software and autonomous agents risk causing chaos. Think Agent Smith from _The Matrix_, running rampant in the real world.
This shift also points to a broader change: the diminishing importance of certain programming languages. In the future, most software creators won’t care which language they’re using. They’ll simply rely on AI models to choose the most suitable one for the task. That’s a fundamental shift.
We’ve seen this before. Once upon a time, people wrote machine code. Today, almost no one does. The same abstraction trend is now playing out with programming languages.
All in all, it appears that the ship has sailed with vibe coding—it may be hard to turn back the clock on something that is democratizing something previously so inaccessible that you needed hundreds of thousands of dollars to build it.
Yet amid the hype, the ramifications on both the engineering labor market and the safety of our digital projects probably shouldn’t be ignored. For example, a few days ago, a writer who used OpenAI’s coding tool Codex created a centralized video surveillance feed of cities around the world. While the writer herself was not malicious in her intent, stories like hers—as well as experiments from other white-hat hackers—are illuminating vulnerabilities.
While vibe coding might certainly be a future income stream for Gen Alpha coders who have innovative and scrappy ideas, there are significant questions around how it will erode middle-management knowledge worker roles today, and also how it might compromise stability and security of digital products. Perhaps, while the ship has sailed, there will be real and necessary jobs opening up when it comes to the ship’s maintenance, as well as tighter protocols on how many people and who can board the ship.
Some other things to pay attention to:
Teens allege Musk’s Grok chatbot made sexual images of them as minors
This CEO warns that Democratic voters are most at risk from automation
His Father Lost His Life’s Savings in a Scam. A Fake Lawyer Offered to Help.
AI companies want to harvest improv actors’ skills to train AI on human emotion
See you next week!