According to Forbes, Henry Shi created the “Top 10 Lean AI Native Companies Leaderboard” in March 2024 to track startups racing toward $1 billion valuations with tiny teams. His list now includes 44 companies that must have $5 million in annual recurring revenue, be under five years old, and employ fewer than 30 people. Topping the current revenue list is Telegram with $1 billion annually and just 30 employees, giving it a staggering $1 billion valuation per employee. Second is Midjourney with $500 million in revenue and 40 employees. Meanwhile, Sam Altman and his CEO friends have created a betting pool to predict when a solopreneur will first hit the $1 billion mark using AI agents.
The leaderboard effect
Here’s the thing about leaderboards – they create competition. And Shi’s Lean AI Leaderboard has apparently become a magnet for ambitious founders who want to prove they can build massive companies with minimal teams. I think there’s something deeply compelling about this idea in the age of AI. We’re basically watching the natural extension of the million-dollar solopreneur trend that started gaining real traction around 2013.
But let’s put these numbers in perspective. The U.S. Census Bureau shows there were 117,060 solopreneurs and partnerships hitting $1 million in revenue in 2023. That number more than doubled from 2021 to 2022. Still, when you consider there are over 30 million nonemployer businesses in the U.S. with average revenue around $57,611, million-dollar solo businesses remain the Olympic athletes of entrepreneurship. Billion-dollar ones would be… well, unicorns among unicorns.
The AI management reality
Now, here’s where it gets really interesting. At a Davos panel on the “One-Person Enterprise,” You.com CEO Richard Socher dropped some truth bombs about what this actually means. He suggested we’re all going to become managers of AI eventually. But running an entire unicorn with just AI agents? That might be further out than people think.
Steve King from Emergent Research, who’s participating in the NYPL panel series on this topic, basically says he has a hard time seeing one person running the whole show. Accounting, finance, marketing, sales – AI can handle agentic work, but project management across all these functions? That’s still a human-heavy lift. His take: one or two people plus freelance help seems more realistic than a true solo operation.
Celebrity versus AI
King makes another fascinating point – the first billion-dollar solopreneur might not come from the AI world at all. He points to celebrity entrepreneurs like Steph Curry, whose company Thirty Ink reportedly earned $174 million in 2024 through sponsorships, speeches, and documentaries. With that kind of leverage, building a billion-dollar business becomes more about personal brand than AI automation.
But data scientist Trent Fowler, another NYPL panel participant, throws some cold water on the AI solopreneur hype. He notes that improvement in AI models is entirely unpredictable. More importantly, he raises the security question – even if a top 1% programmer and business genius creates a billion-dollar platform, how long before security flaws cause it to “go up in flames”?
Funding landscape shift
Here’s what could really change the game though. If AI agents can handle the work without needing salaries, the traditional funding model gets turned upside down. Salaries and benefits are typically the biggest costs in scaling. Remove that, and suddenly you don’t need venture capital to grow quickly. As Shi puts it, “A lot of these AI companies now can get to revenue quickly. You don’t really need venture.”
So where does this leave us? The conversation has clearly moved from whether million-dollar solo businesses are possible to whether billion-dollar ones are imminent. With events like the NYPL panel series bringing together experts and the Altman betting pool adding some Silicon Valley stakes, the race is officially on. But whether the winner will be an AI-powered coder or a celebrity influencer? That’s the billion-dollar question nobody can answer yet.
