Chasing Moats After AI: The Overlooked Goldmine of Niche Pain Points

Pieter Levels is back at it again, this time scrambling to chase remaining moats as AI companies suck up everything else. It's a familiar scramble for anyone who's been following his journey from remote work advocate to AI app mogul. And his list of moats—proprietary data, network effects, distribution, physical components, capital, and regulation—is a solid starting point. But after reading his piece, I couldn't help but think he's missing the biggest moat of all: the long tail of specific, painful problems that are too small for AI giants to care about but too painful for the people experiencing them to ignore.

Levels talks about proprietary data as a moat, and he's built a paid API for nomads.com city data. That's smart. But proprietary data doesn't have to be about places or reviews. It can be about pain. At PainSignal, we track problems people are actually complaining about—the kind of problems that make them willing to pull out their wallets. We've collected 26,388 problems across 89 industries. These aren't vague "AI will disrupt everything" hand-waves. They're specific, like "Traditional bank financing is too slow for real estate investors" with a severity score of 5 out of 5. That's a pain point with a built-in moat: if you solve it, you own the solution.

Levels also mentions network effects with his free community. Network effects are great, but they're not the only game in town. Look at real estate: we track 950 problems in that industry alone. One of them is "Real estate agents are being pressured to promote Zillow Home Loans under threat of lead shutoff, raising RESPA compliance." That's a regulatory moat, the very thing Levels admits he's bad at and avoids. But someone could build a compliance tool that helps agents navigate RESPA without getting their leads cut off. That's a moat of trust and expertise, not just network effects.

He's right about distribution being a moat—his email list of 187,279 subscribers is impressive, though the number in his footer says 184,609, so there's a small discrepancy. But distribution alone doesn't guarantee you're solving the right problem. The real moat is knowing exactly which problems to solve. We track 11,910 app ideas generated from those problems, and many have explicit willingness to pay. For example, "Real estate agents lack a repeatable method to generate leads from AI tools" has a severity of 4/5 and an opportunity score of 62/100. That's a distribution problem waiting to be solved, and it's not on Levels' list.

Physical components—hardware, locations, IRL experiences—are another moat he's chasing. He's building a Pieter.com retro PC mini and promoting meetups. That's a hedge against AI, but it's not the only physical moat. Think about all the physical-world problems that still need software solutions. We see pain points in fitness, travel, and hospitality that have a physical component but are solvable with software. For instance, hotel booking sites suffer from rating inflation, as Levels notes with hotelist.com. Our data shows similar pain points in travel and hospitality, like "Traditional bank financing is too slow" for hotel renovations or acquisitions. That's a physical asset problem with a fintech solution.

Capital is a moat he's leaning into with investments in AI companies, bonds, and real estate. That's a smart hedge. But startup investing is another game entirely. He wonders what startups are still valuable to invest in when Big AI and Big Tech suck up everything. The answer is in the data: startups that solve the long tail of operational pain points. The last YC batch might be all "Cursor for {industry}" harnesses and hardware, as he says, but that's just the visible tip. Beneath the surface, there are thousands of problems like "Real estate agents have an inoperative license" with severity 5/5. That's a compliance issue that needs a simple, automated fix. Not glamorous, but valuable.

Regulatory moats are the one thing Levels admits he's bad at. But our data shows they're surprisingly powerful. Problems like RESPA compliance and licensing issues have severity scores of 5/5 and opportunity scores above 57/100. That means people are desperate for solutions and willing to pay. A startup that can navigate the regulatory maze and offer a turnkey compliance tool could build a moat that's nearly impossible for AI to erode overnight. Regulation isn't going away; if anything, it's getting more complex. That complexity is a feature for those who can master it.

So here's the thing: Levels is right that AI is sucking up everything, but only everything that's visible and general. The long tail of specific, painful, often unsexy problems remains. And that's where the real moats are. At PainSignal, we've turned this data into a map. We track 26,388 problems across 89 industries, each with severity and opportunity scores. We see where the pain is highest and where the willingness to pay is strongest. We see problems that big AI companies will never touch because they're too niche, too regulated, or too human.

For indie hackers and vibe coders, this is the playbook: stop trying to compete with AI on the general stuff, and instead go deep on a specific pain point. Use AI as a tool, not a competitor. For seed investors, the data shows where the next breakout startups might hide—not in the AI hype cycle, but in the operational trenches of real estate, hospitality, healthcare, and beyond. The moat isn't just proprietary data; it's proprietary insight into what actually hurts.

Levels is chasing moats, but he's looking through the wrong end of the telescope. The moats aren't disappearing; they're just getting more specific. And the map to them is in the complaints, the forum posts, the one-star reviews, and the desperate tweets. We've been collecting those signals. Now we're sharing them.