In early 2025, a 28-year-old Israeli developer named Maor Shlomo sat down and started building a no-code platform. No co-founder. No team. No employees. Just him, a laptop, and a growing collection of AI tools. Six months later, he’d built Base44 — a fully functional platform that let non-technical users build web applications with natural language — entirely by himself. Then Wix bought it for $80 million.

Let me say that again, because the number still surprises me every time I type it. One person. Six months. Eighty million dollars.

This is the story that the “one-person unicorn” crowd has been waiting for. Not a unicorn exactly — $80 million isn’t a billion — but close enough to make everyone sit up and pay attention. If a solo developer can build an $80 million product in half a year, what’s stopping the next one from hitting ten figures?

Quite a lot, actually. But we’ll get there.

The betting pool

Sam Altman dropped a line in early 2024 that’s been ricocheting around Silicon Valley ever since:

“In my little groupchat with my tech CEO friends there’s this betting pool for the first year that there is a one-person billion dollar company.”

He didn’t say if. He said what year. The framing matters — for Altman, the one-person unicorn isn’t a question of possibility, it’s a question of timing. Dario Amodei, CEO of Anthropic (the company behind Claude), went further, putting the odds at 70-80% that it could happen by 2026.

As of February 2026, nobody’s claimed the prize.

But the near-misses are piling up fast enough to make you wonder.

Timeline of shrinking team sizes building increasingly valuable companies

The proof points keep coming

Here’s where things get genuinely interesting. Forget the one-person unicorn for a moment and look at what tiny teams are actually pulling off:

Cursor hit $1 billion in annual recurring revenue in just 17 months — the fastest any B2B company has ever reached that milestone. When they crossed $100 million ARR, they had roughly 20 employees. Their marketing budget to get to $100M? Zero dollars. The product spread entirely through developers telling other developers.

Lovable, a Swedish AI coding startup, crossed $100 million ARR in eight months with about 45 employees and zero paid acquisition. Eight months. I’ve seen enterprise sales cycles that take longer than that.

Bolt went from launch to $20 million ARR in two months. Two months. With 15 people.

Midjourney was generating $200 million in annual revenue with a team of about 10 people. That’s $20 million per employee — a ratio that would’ve been science fiction five years ago.

None of these are one-person companies. But they’re all radically small teams producing outcomes that used to require hundreds of engineers, entire marketing departments, and floors of office space. The trend line is unmistakable: the minimum viable team keeps shrinking.

The numbers behind the shrinkage

It’s not just anecdotes. The data tells the same story at scale.

Seed-stage startup teams averaged 10.3 people in 2021. By 2025, that number had dropped to 6.2 — a 40% reduction in just four years. Meanwhile, solo founders went from 22.2% of all new startups in 2015 to 38% in 2024, according to Carta’s data. Nearly four in ten new startups now have a single founder. That’s not a trend. That’s a structural shift.

Garry Tan, president of Y Combinator, has been unusually blunt about what he’s seeing in the latest batches:

“You don’t need a team of 50 or 100 engineers.”

YC’s most recent batch has been their fastest-growing in the fund’s history, and the companies driving that growth are smaller than ever. What once required 20 developers can now be done by 2 or 3 who know how to leverage AI effectively.

(I spoke with a friend who recently went through YC. He told me his two-person team shipped more features in their first month than his previous ten-person startup shipped in a quarter. He wasn’t bragging — he seemed almost unsettled by it, like the rules he’d spent years learning had been quietly rewritten.)

The Shopify mandate

If you want to understand how seriously the industry is taking this shift, look at Shopify. In April 2025, CEO Tobi Lutke sent an internal memo that leaked almost immediately — as these things do — declaring that teams must prove AI can’t do the work before asking for additional headcount. Not “consider using AI first.” Not “explore AI options.” Prove it can’t do it. AI proficiency was added to performance reviews.

The memo was polarizing. Some people called it visionary. Others called it a recipe for burnout and a way to justify never hiring. But what happened next was more telling than the memo itself: the rest of the industry quietly started doing the same thing. Not all of them were as explicit about it, but the pattern was clear in the hiring data. Tech job postings in 2025 dropped significantly while output — by most accounts — continued to rise.

The message from leadership across the industry is converging: small teams augmented by AI are the new default. Large teams are now something you have to justify, not the other way around.

So where’s the unicorn?

Here’s the honest part. The part the hype cycle doesn’t like to dwell on.

Despite all the proof points, despite the betting pools and the 70-80% odds and the shrinking team sizes — as of right now, the one-person billion-dollar company doesn’t exist. Not one. Altman’s groupchat is still waiting.

And there are real reasons to wonder if it’s possible at all, at least in the way people imagine it.

Building isn’t maintaining. Maor Shlomo built Base44 alone and sold it to Wix. But running a billion-dollar company means customer support, legal compliance, financial operations, partnerships, sales — the mountain of non-coding work that scales with revenue. AI can help with some of this. It can’t do all of it. Not yet.

Solo founders get less money. Here’s a stat that complicates the narrative: despite representing 30% of startups, solo-founded companies received only 14.7% of total venture capital cash, according to The VC Corner’s analysis. Investors still have a strong preference for teams. They want to see complementary skill sets, resilience against burnout, and someone to argue with the founder when they’re wrong. A solo founder, no matter how talented, is a single point of failure.

Long-term scalability is unproven. The companies I listed above — Cursor, Lovable, Bolt — are all very young. They’ve built impressive products fast. But can AI-coded products be maintained at scale over years? Can a 15-person team handle the complexity of a mature product with thousands of enterprise customers, each with their own edge cases? We genuinely don’t know yet. The experiment is still running.

The one-person unicorn is venture capital’s Bigfoot — lots of excitement, lots of blurry photos, no confirmed specimen.

The optimist’s case (and I’m making it)

Despite all the caveats, I think the directional trend matters more than whether someone technically hits the billion-dollar mark solo. Here’s why.

The floor for what a small team can build has risen dramatically. Five years ago, a solo developer could build a side project. Today, a solo developer can build a company worth tens of millions. The gap between “side project” and “real business” has collapsed. That’s meaningful even if no one person ever hits a billion.

And the democratization angle is real. Building software used to require either deep technical expertise or enough money to hire people who had it. AI is eroding both barriers simultaneously. The Bessemer Venture Partners’ State of AI report for 2025 highlighted that the cost of building an MVP has dropped by roughly 90% compared to five years ago. That means more people can try. More experiments get run. More weird, niche, wouldn’t-have-been-funded ideas get built.

Is every one of those experiments going to succeed? Of course not. Most will fail, same as always. But the volume of attempts is increasing, and in startup math, volume matters. The more shots on goal, the more goals.

The Y Combinator data backs this up. Their recent batches aren’t just smaller — they’re more diverse in terms of founder backgrounds. People who wouldn’t have been able to build a tech startup three years ago are now doing it, because AI has lowered the technical barrier enough that domain expertise matters more than engineering headcount.

What this actually means for builders

If you’re a developer reading this, the takeaway isn’t “go be a solo founder.” Most people shouldn’t. Running a company — even a tiny one — is a completely different skill set from building software, and AI doesn’t change that.

The takeaway is subtler: the leverage available to small teams has increased by an order of magnitude, and the ceiling keeps rising. Whether you’re building a startup, freelancing, or working inside a larger company, the amount of impact you can have as an individual or a small group is larger than it’s ever been.

That’s genuinely exciting. Not in the breathless, Silicon Valley, “disruption” way. In the practical, tangible, “I can actually build this thing I’ve been thinking about” way.

(I’ve been noodling on a project for two years that I kept shelving because it would’ve required at least three months of full-time work to get to an MVP. Last month I prototyped the core functionality in a weekend. It’s not done. It’s not good enough to ship. But it exists, and that changes the calculation entirely.)

The one-person unicorn might never arrive. But the one-person company that generates real revenue, solves real problems, and competes with teams ten times its size? That’s already here. And it’s going to become increasingly common.

Altman’s betting pool might stay open for a while. But in the meantime, the rest of us are building.

Next in this series: Part 7 looks at the other side of this coin — what happens to the developers who don’t become founders. The job market data is more complicated than either the optimists or pessimists want to admit.

Sources