Right now, somewhere in a cafe in Lisbon or Bangkok or Tokyo, a Dutch man in a black tank top is building something that will break the internet in three weeks. His name is Pieter Levels. He has no employees. He has no investors. He has PHP, a cron job list 180 entries deep, and the kind of focus that looks a lot like stubbornness from the outside.
By 2026, Levels pulls in over $3 million a year across a portfolio of products - a digital nomad community, a remote job board, an AI photo generator, an AI interior design tool, and a multiplayer flight simulator built in a single afternoon. He does not delegate. He does not fundraise. He does not take Zoom calls. He does not, pointedly and deliberately, use React.
The flight simulator is probably the most instructive. In February 2025, Levels had never written a single line of game code. He opened Cursor AI, described what he wanted, and three hours later fly.pieter.com was live. Within days, Elon Musk tweeted "Wow, this is cool. AI gaming will be massive." Within three weeks, it had 89,000 players and $1 million in annualized revenue. The game dev industry, which employs hundreds of thousands of people and spends years shipping titles, had just been casually lapped by a man who makes his living by working very fast and caring very little about how things are supposed to be done.
I was afraid of failing in public. So I decided to do it repeatedly until I wasn't afraid.Pieter Levels
The Spreadsheet That Started Everything
The origin story most people know: 2014, Chiang Mai internet cafe. Pieter Levels is a digital nomad trying to figure out where to go next. He opens Google Sheets, types up a list of cities with data on cost of living, internet speed, and temperature, and tweets it. Within 24 hours, 100 people have added their own data. The spreadsheet hits Reddit and Hacker News simultaneously. He rebuilds it as a proper website - layout copied directly from Product Hunt, code written in an afternoon. NomadList launches and immediately goes to the top of both Product Hunt and Hacker News at the same time, an accident nobody planned and almost nobody has replicated since.
What the spreadsheet story misses is everything that came before it. At 12, Levels tried to set up an online business and had his father sign a Worldpay merchant agreement with a potential $100 million liability. He never made a sale. As a teenager he produced techno and drum and bass under multiple aliases - music that went nowhere commercially but trained the instinct to make things and release them. In 2007, he started teaching music production on YouTube, before anyone called that content creation, and grew a channel to 139,000 subscribers earning $8,000 a month. Then in 2012, overnight, YouTube's algorithm changed and the income evaporated. He had a panic attack he later attributed, with characteristic precision, to drinking six or more cups of coffee per day.
To afford his one-way ticket to Bangkok in 2013, Levels sold nearly everything he owned: PC disassembled and sold for parts on eBay, DSLR camera, musical equipment. He raised about $3,000. He left Amsterdam with a MacBook, an external hard drive, four t-shirts, two pairs of pants, six pairs of socks, and six pairs of underwear. That was it. He wrote the entire experience up in a post called "Reset Your Life" that got shared hundreds of thousands of times and quietly convinced a generation of burnt-out people that radical reinvention was an actual option, not just something you say.
12 Startups in 12 Months
The reset led to the challenge: one product launched per month for 12 months. The explicit goal was to cure himself of the perfectionism that had kept him from shipping. "Launch at 70% completion rather than pursuing perfection" became his operating principle - not a productivity hack, but a diagnosis of his own failure mode and a prescription for fixing it.
Most of the 12 startups failed. Play My Inbox: gone. Tubelytics: gone. GifBook: gone. Go Fucking Do It - an accountability app that charged your credit card if you missed a self-imposed deadline - still has a cult following. Remote OK, launched as part of the challenge, became the most widely used remote job board on the internet. And NomadList, the accidental spreadsheet, became the defining digital product of the remote work revolution that would take the rest of the world another six years to catch up to.
Over time, the total count of launched projects grew past 70. By his own math, four of them made real money. He treats this 5% hit rate not as evidence of failure but as evidence of the system working correctly. The 66 that failed were the cost of finding the four that mattered.
Only 4 out of 70+ projects I ever did made money and grew. Make what you know. Most startups don't need to scale. They need to be profitable.Pieter Levels
The AI Pivot Nobody Saw Coming
In September 2022, Levels started experimenting with Stable Diffusion. He noticed it could do something interesting with rooms. He built InteriorAI - upload a photo of your space, get an AI redesign back - in roughly a weekend. It hit $50,000 a month in revenue without a marketing budget or a team. The lesson, which he tweeted loudly and repeatedly, was that Stable Diffusion and its successors had collapsed the cost of building AI products to something close to zero for a competent solo developer.
In February 2023, PhotoAI launched. The concept: upload photos of yourself, get AI-generated headshots, portraits, and lifestyle photos back in any setting. The product made $100,000 in its first 10 days. Within 18 months it was generating $100,000+ per month in revenue and had become the anchor of his entire business. He paid $40,000 for the photoai.com domain - at the time his single largest capital expenditure - and considered it an easy decision.
The pattern in both cases is the same one that produced NomadList: identify a problem he personally has, build the minimum version fast, post about it publicly, let early users shape the product. No market research. No user discovery sessions. No product manager. No roadmap deck. Just a person who needs something, builds it, and finds out by launching whether anyone else needs it too.
The Stack Question
Levels runs his entire portfolio - products earning tens of thousands of dollars per month each - on vanilla PHP, jQuery, MySQL, and SQLite. He runs 180 cron jobs to automate operations. He has written publicly and repeatedly about his refusal to adopt modern JavaScript frameworks, and equally publicly about the judgment he gets for it from people who find React more interesting than outcomes.
"PHP just stays the same and works," he has said, with the confidence of someone who has never had to explain a dependency conflict to a user who just wanted to pay for a subscription. The pragmatism is structural: a solo developer maintaining six products cannot afford the cognitive overhead of keeping up with the JavaScript ecosystem. The old boring tools run quietly and require no maintenance attention. That frees capacity for actually building things.
He holds a business degree from Rotterdam School of Management and studied music production at the Conservatorium van Amsterdam. He has no formal computer science credentials. He learned to code as a child on MS-DOS Batch and Windows 3.1, continued on his own through his teenage years, and built his entire $3 million operation on that self-taught foundation. He has said repeatedly that engineers who want to start companies should learn business rather than the reverse, and that the combination of technical skill and commercial sense is rarer than either alone.
The Movement He Didn't Plan to Lead
When NomadList launched in 2014, "digital nomad" was a phrase that required explanation. Levels spent the next several years not just building the platform but defining the category - creating a Slack community that grew to 40,000 members (now moderated partly by GPT APIs), writing about what it actually felt like to live and work nomadically, including the parts nobody talked about: the depression, the loneliness in hotel rooms, the way that total freedom can produce total paralysis.
He was the first person ever interviewed on the Indie Hackers podcast in 2016, at a moment when "indie hacking" was not yet a recognizable job description. The transcript of that interview reads, in retrospect, like a manifesto for an entire subculture that took five more years to fully emerge. He won Product Hunt's Maker of the Year award in 2018, the year he published MAKE: Bootstrapper's Handbook, a guide to building and selling products solo that has sold over 10,000 copies.
In 2024, Levels co-founded EU/acc - European Accelerationism - as what he described as "a fork of e/acc for Europe." The movement's core positions: reduce EU tech regulation, eliminate cookie consent banners, create a streamlined EU Inc. entity for startups, and reform tax policy to favor founders. The crowdsourced manifesto that emerged apparently caught enough attention in Brussels that European Commission planners began citing it in discussions about scaling back GDPR scope and softening the AI Act. A manifesto on Twitter turning into EU policy discussions is, in Pieter Levels terms, just another Tuesday.
The Vibe Coding Era
The flight simulator is the purest expression of what Levels has always been doing, accelerated by AI. The premise of vibe coding - describe what you want to an AI coding assistant, accept what it gives you, iterate - is philosophically identical to the "ship at 70%, learn by doing" approach he has practiced since 2014. The AI just collapsed the timeline from months to hours.
In 2025, he organized the Vibe Code Game Jam, posting $40,000 in prizes and attracting 1,170 game submissions. In 2026 he is running Cursor Vibe Jam 2026 with a $35,000 prize pool, requiring that 90% or more of the code be AI-generated. He is not positioning himself as a curator of these events. He is, in the most literal sense, the person who proved the concept by building a game in a workday and watching Elon Musk tweet about it.
The Lex Fridman conversation in August 2024 - four hours, published to an audience of millions - provided the sharpest outside window into how Levels thinks. He talked about programming languages the way other people talk about tools: preference without ideology. He talked about PhotoAI and its training data challenges with the same flat candor he brings to everything. He talked about robotics and the future of work with the conviction of someone who spent a decade watching everyone catch up to ideas he had already built products around.
What the Fridman episode made clear, to anyone watching, is that Levels is not performing the indie hacker persona. He is genuinely just a Dutch programmer who moved to Asia with four shirts and built several companies by doing things himself, fast, in public, and without permission. The tank top is not a brand choice. It is probably just comfortable.