On Friday, I read Kevin Roose’s New York Times piece about how anyone can be a coder with just a single idea and a few well-constructed prompts to no-code AI tools. He explores the in-vogue concept of vibecoding, a newish phenomenon that's been taking the tech industry by storm (and leading to lots of wild demos on social media about the AI doing the work for you).
In Kevin's piece, he also shares some of the "software for one" tools he's built for himself without coding, largely micro-apps and bespoke productivity tools, like a meal planner that suggests recipes based on what’s in his fridge.
I appreciate the growing awareness around this idea, but I think Kevin’s column only scratches the surface. There’s a much bigger story here—one that goes beyond personal productivity hacks and into the broader implications of how AI is reshaping who gets to build, what gets built, and how we think about software itself.
I actually emailed Kevin on Friday with a few thoughts on what he might have missed. Since I haven’t heard back, I figured I’d share them here instead.
What He Missed:
1. Vibecoding is transforming who can build things online.
Why it Matters:
There’s plenty of discussion about how these new AI-powered tools make it easier for anyone to build software, but we haven’t fully explored what happens when people outside the traditional tech sphere—those who look different, think differently, or come from entirely different backgrounds—start creating with them.
The fact that women still make up only about 25% of software developers—and that women-founded startups receive less than 3% of venture capital funding—is, frankly, an embarrassment. Why aren’t we treating no-code AI as a chance to finally—maybe—correct some of the damage from the past two decades?
It’s no surprise when someone already embedded in tech picks up a new hobby using the latest tools. That pattern repeats with every new tech wave. But the more interesting question is: who else is picking up these tools? What are they building? Or who else might? And how might people who have never touched a command line—who don’t even have the vocabulary for it—start conceptualizing entirely new kinds of problems? While Kevin shared a few examples in his article, I think he missed a big opportunity to invite many more diverse voices (particularly, voices of women and other under-represented groups) into this new narrative.
It's anyone's guess about what will happen, so here's mine. We’ll start seeing less “software for one” and more “software for the overlooked.” Instead of just productivity hacks and meal planners, we’ll see tools designed for niche, marginalized, or historically ignored communities—built by the very people who understand their needs firsthand. By the way, when that happens, vibecoding won’t just be about vibes. It will be a quiet but powerful shift, reshaping access, autonomy, and problem-solving on a much larger scale. Kind of a big deal.
What He Missed:
2. Vibecoding is for More than Just Hobby Projects
Why It Matters:
There's been a lot of talk about how apps like Cursor, Bolt, Lovable, and Vercel can get v1 prototypes out the door. Naturally, good AI consultants are completely overbooked, and we're already starting to see no-code AI trainings popping up, with everyday hobbyists taking a swing at the same continuing education market as major educational institutions.
But the real story isn’t just about accessibility. It’s about what people are doing with this newfound superpower. Kevin’s piece focused on the novelty factor—fun, niche projects—but missed the bigger picture: people aren’t just tinkering, they’re launching real businesses. Anyone following people like Greg Isenberg or Ben Lang knows this isn’t a hypothetical “What if?”—it’s already happening.
As someone doubling down on my own AI-native, AI-built startup, I am seeing and feeling this shift firsthand. Stopping the narrative at “vibes” is a huge oversight—because what’s happening right now is much bigger than that.
What He Missed
3. This New Era of Building Requires a New Way of Thinking
Why it Matters:
There’s a distinct mindset to coding—part vibes, part vocabulary, and part holistic, big-picture thinking. The old-school image of a siloed technologist, focused only on the code, is fading fast. Because today, there’s the code… and then there’s everything else. And the more coding gets automated, the more important that everything else becomes.
This shift creates a new kind of power player: Someone who can code a little bit but also brings skills from outside of tech—design, storytelling, systems thinking, business strategy. Today’s generalist is tomorrow’s superthreat.
When coding becomes truly democratized (which, arguably, has already happened), building software isn’t the tricky part anymore. And if coding isn’t the hard part, then all the so-called backburner business skills—communication, creativity, domain expertise—become the new center of gravity. After all, in a world where anyone can code, everyone will also need to be something else.
Maybe that’s a gross oversimplification of what’s coming. But my point is this: The narrative is changing. This isn’t just about hobby projects or a fun, insider-only phenomenon. AI-driven coding is a powerful shift that’s giving people agency, expanding access, and making skills outside of software development more essential than ever. And that’s a much bigger deal than just “vibes.”
