Why Women Are the Real Vibe Coders Now
How I Learned to Stop Worrying and Love The “Frankenstein” Known As Vibe Coding
You know that moment in the book Frankenstein when Victor realizes his creation has escaped into the world, and suddenly everyone knows the Monster better than they know him?
That’s exactly what happened to Andrej Karpathy in 2025.
The man coined “vibe coding,” which is this beautiful, chaotic idea that we could just talk to computers like they were our creative besties, and it took off like wildfire. Everyone was doing it. Everyone was talking about it. The tech bros, the wannabe founders, the “I have an app idea” guys at parties.
And then, in the ultimate plot twist that would make Mary Shelley herself slow clap, Karpathy announced he hand-coded his new project. Didn’t vibe code it at all. Said the AI was “net unhelpful.”
The internet lost its mind. “Vibe coding is DEAD,” they declared on Blind, the tech gossip forum where hope goes to die. February to October 2025: the official death dates of a practice that barely got to celebrate its first birthday.
But here’s what actually happened, and why it matters especially if you’re a woman trying to break into tech, build something meaningful, or just trying to understand why everyone’s suddenly writing eulogies for something that never actually died.
The Creation Myth (Every Good Monster Story Starts Here)
Let me take you back to when this whole thing started, because context is everything.
Andrej Karpathy, former Tesla AI director, OpenAI founding member, basically tech royalty, gets on stage at Y Combinator and describes this wild new paradigm. He says we can now program computers in “our native language of English.” Not Python. Not JavaScript. Just... talking. Like you’re explaining your idea to a friend who happens to be really good at making it real.
The vibes (pun!) were immaculate. He demonstrated by building an iOS app in Swift, a language he didn’t even know, just by describing what he wanted. A restaurant menu app. Boom. Weekend project, manifested.
This tweet became what he called “a major meme.” And honestly? It captured something true. That feeling that the barrier between “I have an idea” and “I made a thing” had just... dissolved. Like, suddenly you didn’t need four years of computer science and a GitHub graveyard of abandoned tutorials to build something.
For women getting into technology like myself, this was revolutionary. Because here’s what nobody talks about: we’ve been conditioned to believe we need to know everything before we’re allowed to try anything. That we need to be “technical enough” before we can call ourselves developers. That if we can’t reverse a binary tree on a whiteboard while reciting the entire history of Unix, we don’t belong.
Vibe coding said: screw that. Just describe what you want.
When Your Monster Becomes a Marketing Campaign
The hype cycle that followed was chef’s kiss peak 2020s tech nonsense.
Platforms like Lovable marketed this idea that anyone, literally anyone, could “build an app in minutes!” The promise was seductive. The reality?
By 2025, we started seeing the receipts. Academic papers talked about a “speed-quality trade-off paradox”—aka your code was fast but trash. 404 Media reported on AI-generated code that literally wiped out entire company databases. Like, full scorched-earth data apocalypse because someone trusted the vibes too much.
Even Bain & Company, which you’d think would be cheerleading anything that promises to make software engineers more “productive,” was like “yeah, the savings have been unremarkable.”
Then came the Karpathy betrayal. His Nanochat project? Hand-coded. All of it. He said AI agents were “net unhelpful” and his codebase was “too far off the data distribution.”
The collective rage was real, kinda like when in Dragonball Z when Frieza killed Krillin and Goku went nuts? Except instead of righteous fury, it was a bunch of tech people feeling personally victimized by a methodology shift.
But here’s what I keep thinking about: Victor Frankenstein also abandoned his creation. He made this thing, saw it had problems, and peaced out. The Monster wandered around trying to figure out what it was supposed to be, learning, evolving, getting better at existing in the world.
Karpathy didn’t kill vibe coding. He just admitted that his creation had grown beyond him in ways he didn’t expect. The Monster was loose, and it was doing its own thing now.
The GoPro in Your Closet (We All Have One)
Okay, so there’s this YouTube video by Theo that absolutely nailed what happened next, and I cannot stop thinking about it.
He compared vibe coding platforms to GoPro cameras. Stay with me.
GoPro didn’t just sell a camera. They sold a fantasy. You, shredding down a mountain. You, surfing perfect waves. You, as the protagonist of your own action movie. Everyone bought one. Everyone used it twice. Everyone’s GoPro is now in a closet somewhere, next to the bread maker and the exercise bike.
Vibe coding platforms sold the same dream: You, as a software developer. You, building the next big app. You, disrupting industries from your bedroom.
The traffic data tells the story. Massive surge when these tools launched. Then a steep drop. The novelty tourists showed up, realized building actual products still requires sustained effort and problem-solving, and left.
But, and this is crucial, Lovable’s founder said something interesting. While web traffic fell, paying customers are growing.
The experimenters churned. The serious users stayed.
Which brings me to the part nobody’s really talking about: who are the serious users?
Who stuck around after the hype died and the bros moved on to the next shiny thing?
The Women Who Never Left (Because We Never Got to Stay)
Here’s where I need to get real with you.
When the tech discourse talks about “vibe coding’s death,” what they’re really mourning is the loss of a particular kind of user: the casual tinkerer. The weekend warrior. The guy who wanted to spin up a SaaS on a Saturday afternoon and retire by Tuesday.
But there’s a whole other population that never shows up in these conversations.
Women who are using these tools not as a magic wand, but as a force multiplier. Women who were never looking for a shortcut, because we’ve never been afforded shortcuts, but for a way in.
Think about it. The gender gap in computer science is horrifying. Women make up about 20% of CS graduates. The reasons are systemic, cultural, confidence-related, pipeline-related, and a thousand other “-related” that we could write dissertations about.
Vibe coding didn’t fix that. But it did something else.
It gave women who were told they “weren’t technical enough” a way to start building. To prototype. To prove to themselves that they could translate ideas into reality without first completing some arbitrary gauntlet of technical hazing.
I think about the women I met, read about and interacted with on Substack, women like
and and and . They’re not trying to replace engineers. They’re designers who want to understand implementation. Product managers who want to prototype faster. Founders who need to build MVPs without burning through their runway hiring a dev shop that might not understand their vision.They’re using AI tools the way Karpathy actually uses them in his professional work with what he calls the “autonomy slider.” Not as a replacement for expertise, but as an extension of it.
The Autonomy Slider (Or: How to Ride Your Monster)
Here’s what Karpathy actually does with AI, according to his own breakdown of building Nanochat:
Low autonomy (primary mode): AI as super-powered autocomplete. He’s the architect, pointing the AI exactly where to go, typing the first few characters, getting high-bandwidth completions. Humans are very much in control.
Medium autonomy (selective use): For bounded tasks, boilerplate, or languages he’s less familiar with (he mentions Rust), he lets AI take a bit more lead. This is where “a bit of vibe coding” lives—increasing accessibility to new paradigms.
High autonomy (mostly rejected): The full “Hi, please implement this feature” agentic approach. This is what didn’t work for his core project. It bloated his code, misunderstood his architecture, and used deprecated APIs.
Notice something? He didn’t abandon the practice. He just grew up about it.
And you know who’s really good at this kind of nuanced tool use? Women. Because we’ve been trained our entire lives to read situations, modulate our approach, and use every resource available to get taken seriously in spaces that weren’t built for us.
The women I see vibe coding aren’t treating AI like a magic wand that’ll make them instant billionaires. They’re treating it like Sailor Moon’s transformation compact. As a tool that amplifies their existing power when they need it, but they still have to show up as themselves and do the actual work.
(And yes, I just compared vibe coding to magical girl transformations. You’re welcome.)
The Gateway Drug (And Why That’s Actually Good)
Theo called vibe coding a “gateway drug,” and Karpathy agreed. He is specifically pointing to kids learning to code through these tools.
But let’s talk about why the “gateway drug” framing matters for women specifically.
Traditional CS education has been a disaster for retaining women. We show up, we’re often the only woman in the room, we’re subjected to subtle (and not-so-subtle) messages that we don’t belong, and we leave. The statistics are brutal.
But vibe coding as an entry point? That’s different.
It’s private. You’re in your room, in your own space, talking to an AI that doesn’t judge you for asking “stupid questions.” You can experiment without the performance anxiety of a classroom where everyone else seems to already know everything.
You can build something, even if it’s janky, and see it work. That first “oh shit, it works” moment? That’s the dopamine hit that keeps you going. That’s what makes you think, “maybe I can do this.”
I did this constantly. Started by vibing out a simple app. It’s rough. It’s got bugs. But it’s theirs, and it exists. Then they want to make it better. So they start learning about state management, or API calls, or why their authentication keeps breaking.
The vibe coding got them in the door. Curiosity and problem-solving keep them there.
This is the gateway drug effect working exactly as intended. Not as a replacement for deeper learning, but as the spark that makes you want to learn deeper.
The Bifurcation (Or: Two Monsters, Two Purposes)
So here’s what actually happened to vibe coding. It didn’t die. It split.
For novices: It’s an educational on-ramp. A prototyping tool. A confidence builder. A way to go from “I have an idea” to “I made a thing” without first needing a CS degree.
For professionals: It’s a sophisticated assistant. An accessibility tool for unfamiliar languages. A boilerplate generator. A force multiplier that lets you move faster in the areas where you already have expertise.
The mistake was never thinking it would be one thing for all people.
The mistake was the hype cycle convincing casual users that they could skip the learning, the iteration, and the problem-solving. Thinking they could skip the actual work of creation.
Victor Frankenstein’s mistake wasn’t creating the Monster. It was abandoning it instead of teaching it how to exist in the world. The Monster wasn’t inherently evil. It just needed guidance, boundaries, and realistic expectations about what it could and couldn’t do.
What Died, What Lives, What’s Coming
The “vibe coding is dead” discourse is mourning the wrong thing.
What died: The novelty. The hype. The fantasy that you could speak an enterprise-grade application into existence without breaking a sweat.
What lives: The practice itself, evolved and bifurcated. The genuine utility for both beginners and experts. The tools are getting better, more specialized, more focused on sustainable creation rather than instant gratification.
What’s coming: A more mature ecosystem. Better UX. Guided support. Integrated testing. Tools that meet users where they actually are, not where marketing campaigns pretend they are.
And most importantly, though the tech press keeps missing this, women continue to use these tools to build things, learn things, and carve out space in an industry that’s never made space for us.
The Monster Speaks (A Conclusion)
There’s a moment near the end of Frankenstein when the Monster gets to tell his side of the story. He’s not the villain everyone made him out to be. He’s just a being trying to exist in a world that was terrified of him, that attacked him on sight, that never gave him a chance to be anything other than a monster.
Vibe coding got the same treatment. It was hyped to heaven, then attacked at the first sign it wasn’t magic, then declared dead when it didn’t live up to impossible promises it never should have made.
But like Shelley’s Monster, it’s still here. I’m still learning. Still evolving. Still helping people create things.
The men who loudly declared it dead have moved on to the next discourse, the next hype cycle, the next thing to be an early adopter of and then publicly abandon.
The women? We’re still here, building.
Because we were never here for the hype, we were here for the access. For the opportunity. For the chance to translate our ideas into reality using whatever tools actually work.
Vibe coding isn’t dead. It just stopped being a trend and became a practice.
And honestly? That’s when it got good.
What’s your relationship with vibe coding? Are you using it? Abandoned it? Never tried it? I’m curious about your stories—drop a comment or hit reply if you’re reading this in email. Let’s talk about our Monsters.
Author Spotlight
We’re so grateful to
for allowing us to share her story here on Code Like A Girl. If you enjoyed this piece, we encourage you to visit her publication and subscribe to support her work!
Join Code Like a Girl on Substack
We publish 3 times a week, bringing you:
Technical deep-dives and tutorials from women and non-binary technologists
Personal stories of resilience, bias, breakthroughs, and growth in tech
Actionable insights on leadership, equity, and the future of work
Since 2016, Code Like a Girl has amplified over 1,100 writers and built a thriving global community of readers. What makes this space different is that you’re not just reading stories, you’re joining a community of women in tech who are navigating the same challenges, asking the same questions, and celebrating the same wins.
Subscribe for free to get our stories, or become a paid subscriber to directly support this work and help us continue amplifying the voices of women and non-binary folks in tech.
 | A guest post by
|