I’ve Been Littering the Internet With Vaporware

In three weeks, I've filled the internet with apps so niche that they may as well be disposable.

Esther Dyson grew up in a house with no television. At the precocious age of eight, she was already publishing her own newspaper, the Dyson Gazette, written out in ballpoint and copied on carbon paper. It was the first sign of a lifelong condition: she could not abide anything that claimed to be more than it was.

She came, admittedly, from a family prone to substance. Her mother, Verena, was a mathematician in Tarski’s logic group at Berkeley. Her father, Freeman Dyson, was the physicist who imagined wrapping a star in a shell of solar panels to drink all of its light, and who won the Templeton Prize for, of all things, progress in religion. A man honored for reaching past what could be proven.

Esther’s gift ran the other way. She had a reporter’s nose for the bluff.

So when she put the word “vaporware” into print in November of 1983, she was not being clever. She was doing what she had always done: naming a kind of dishonesty. “Good ideas,” she wrote, “incompletely implemented.” Products announced into existence that never arrived.

The genre has a hall of fame now. Duke Nukem Forever was promised in 1997 and shipped in 2011, a full fourteen years of a game that lived only in press releases. Apple announced a wireless charging mat called AirPower in 2017 and quietly killed it two years later, having never sold one. The word stuck around because the behavior never stopped, and now it’s become a central tenet of the software iteration hype cycle.

I only came across Esther’s story because I realized I’d been misusing the term myself after exclaiming that I’d spent the last several weeks filling the internet with vaporware of my own. But mine differs from hers in one way that turns out to matter enormously.

Dyson’s vaporware was insubstantial because it didn’t exist.

Mine is insubstantial because it exists all too easily.

I can date it to a single realization only a few weeks old: With Claude Design, I could build an entire web experience in a single file. Every image baked directly into the page, the photograph itself rewritten as Base64 text, thousands of characters the browser reads back as a picture, so there is nothing to load from anywhere else. One file. You drop it onto the internet, and a business exists. Claude Design had barely cleared beta, and I’d already built dozens of fully formed apps, burning through my tokens until the limit cut me off, then waiting for them to reset so I could build more.

A few weeks ago, this still appeared a miracle, so I directed it at the person I most wanted to make something for. I built Amy a private matchmaking network, her own dating portal, a closed garden where she could play the role she plays better than anyone. Forty-eight hours on Base44 and it was live, everything a social network needs, down to the payment rails she’ll never turn on. No Tinder, no Bumble. Just the most exclusive room of high-EQ singles on the planet, her own network of self-healed clients and bootcamp alumni, behind a door only she holds the key to.

One night, I was scrolling Reddit and saw an ad for a headless Shopify build in Lovable. I know the limits of an off-the-shelf Shopify theme better than I’d like to. But now I can build any storefront I want from scratch and wire it directly through the Shopify API. Today I’m shipping fifty-thousand-dollar storefronts in as little as three hours, with product bundling, dynamic pricing, automatic variant selection, all of it conjured in minutes.

Then I built a Netflix. A full streaming interface, chopped up and wired, fully usable and ready to deploy. Liking, following, commenting, bookmarking. A feed for the people you follow and a feed all your own. Profiles. Messaging. Half a day’s work. Client work that, less than a year ago, would have meant paying a premier agency between fifty and a hundred and fifty thousand dollars and spending months negotiating the color of a button. Those jobs just evaporated. They were there three months ago, and now they are the vapor.

My media agency, which now runs across a half dozen domains, each one styled for a different service and a different audience, has tripled its revenue in the last few months. Every proposal now goes out as a single-page interactive site, built in Cursor and delivered as a small piece of software rather than a PDF. I charge less than I used to, and the work is better than it has ever been. So much so that the agencies I’m making irrelevant have started hiring me to white-label the thing that’s making them irrelevant, each one feeding me a little more of the work that used to be theirs. It’s still worth my time, because I have so much more time to give.

I’ve spent my career inside startups. I’ve managed teams that grew from two members to two hundred in eight months. But I have never witnessed scale like this. The scale of one person, alone, delivering twenty times the work at a quality that wasn’t possible before. No standups, no roadmap, no headcount. Just me and the machine, shipping.

Up until a week ago, I was still building decks in PowerPoint, hunting stock photos that looked like I’d taken them, filling graphs with data I’d sweated over for hours. I saved every revision, archived each one so I could look back. No more. Now the deck is a full web app, built in minutes. And when the meeting ends, I close the tab. I don’t archive it. I don’t look back. I’ll make another one tomorrow that’s better, in less time, and that one will be disposable too.

It’s clear who will be sacrificed in the age of AI. It’s just as clear to me who benefits. A man who has never once suffered from a shortage of ideas. A product manager, the role that exists to stand in the gap between the business case and the code. Hand that man the ability to cross the gap alone. The Man is me. The satisfaction is total, it is instant, and it is becoming an addiction. I can feel the restlessness in my hands before breakfast, the same anxiousness I knew as a smoker.

Beneath the revenue, a new kind of cost is coming into focus.

The productivity gain isn’t linear. It compounds. I told Amy recently that I feel my dopamine drop when I don’t launch something. So I built a live dashboard for my gaming clan, the week’s standings scraped from screenshots, because my mind demanded a launch. Every idea I have, I now just make. The distance between the thought and the thing used to be a road so long you could watch the seasons change between builds. Now it’s a doorway. It opens in the morning, and I walk through it by lunch, and the next morning I need another one, a bigger one, just to feel what I felt the day before.

I used to tell Amy I liked cooking dinner and washing the dishes because everything else I worked on had such long lead times. A branding project might not pay off for months. A software platform, the same, if you’re lucky. So I took my small hit of accomplishment from a clean kitchen or a finished meal; the kind of done that arrives the same day you start. The work never gave me that. The dishes did.

Now the work gives it to me a dozen times before dinner, which used to be filet au poivre on a bed of creamed potatoes, asparagus poached while the sauce reduced. Last night, it was Annie’s Mac n’ Cheese and chicken nuggets.

Dyson coined the term “vaporware” to shame people selling things that weren’t there. The strange joke of this moment is that the practice she warned about has become the easy part. Anyone can make the thing now. A Netflix before lunch. A storefront before the coffee’s cold. The promise and the product have fused, and in fusing they’ve cheapened. When a project that took a team six months now takes me six hours, I don’t treasure it. I throw it away. I built it disposable because it was easy, and easy made it worthless to me, even when it works perfectly.

So the old question changes shape. It was never whether I could build it. It’s whether it should exist at all, whether it should exist by my hand, and what all this frictionless, dopamine-fed making is leaving behind. There are hundreds of Git commits and Vercel deploys with my name on them that I never lifted my hand from the mousepad to make. Maybe that’s a body of work. Or maybe I’m just scratching and clawing my way out of the class I was born into, the walls of the well getting smoother every day, the last rough edges sanded off until there’s nothing left to grip. Until then I’ll hold the fn key, speak my next idea into existence, and climb like the rest of us. Before the well closes over.