From Informal Economies to AI: The Opportunities in Front of Us
As an economics major in college, I was drawn to the courses where economic theory bumped up against the messiness of real human behavior. The classes that asked not just how markets work, but how people work - especially people operating outside the systems economists typically model.
The course that made the biggest impression on me was a Development Economics class I took during my study abroad in Australia. It was the first time I encountered the concept of informal economies in any serious way, and it rewired something in how I think.
Informal economies are the economic activity that happens outside formal regulatory frameworks. No contracts, no tax registration, no official market structure. And yet they represent a massive share of global GDP and employ a significant portion of the world's working population. What fascinated me wasn't the absence of structure - it was the presence of ingenuity. Waste pickers in Laos building self-sustaining collection businesses from riverway debris. Informal valets in Baltimore, a city I know well, claiming parking spots with cones and guiding cars in for a few dollars. Traffic in parts of Southeast Asia governed not by signals but by vehicles negotiating through perpendicular flow, bit by bit, until the stream could pass. No infrastructure. No central authority. Just people reading a problem and solving it with what they had.
The City as a Living System
From Development Economics, my intellectual path moved into urban planning, where I encountered a related concept: tactical urbanism. The idea that cities are not just designed top-down, but constantly redesigned bottom-up by the people who use them. No one captured this better than William H. Whyte, a writer and urbanist who spent years studying how people actually used public spaces -- not how architects and planners assumed they would. His Street Life Project produced a 58-minute film of ordinary people doing ordinary things in the courtyard outside a Manhattan office building. People watching. Sitting on ledges. Gravitating toward food carts. Clustering at the edges of fountains. It sounds unremarkable. It was riveting - because Whyte understood that human behavior in space is not random, it is legible, and if you pay close enough attention to it, it tells you everything about what people actually need.
Desire lines are one of the most elegant expressions of this principle - the informal paths worn into grass by pedestrians taking the most direct route, regardless of where the sidewalk was poured. People didn't petition for a new sidewalk. They just walked where they needed to go until the path became undeniable. Tactical urbanism takes that same logic and applies it deliberately: small, low-cost, reversible interventions that test ideas in real space before anyone commits to permanent infrastructure. The human adaptation comes first. The official design follows.
Both of these concepts - informal economies and tactical urbanism - are really about the same thing: people finding the best path to the outcome they need, with or without the resources the system assumes they have. And in both cases, the insight only becomes visible when someone is paying close enough attention to see it.
From Philanthropy to Tech
That thread pulled me into an early career in nonprofits and philanthropy, where I worked alongside social innovators - people building programs and ideas to address problems that markets and governments had failed to solve. The principles were identical to what I had studied. Constrained resources. Real problems. Ingenious solutions. And a belief that if you could just get the right intervention in front of the right community, something would change. But the longer I worked in that space, the more I felt the weight of a structural problem that good intentions alone couldn't solve.
Nonprofit programs - even the ones that worked, even the ones with real evidence behind them - lived and died by funding cycles. I watched program directors spend more time writing grants than delivering the programs those grants were supposed to fund. I watched genuinely effective interventions fail to scale not because the model was wrong, but because the revenue model underneath it couldn't sustain the growth. The impact was real. The infrastructure to compound it wasn't.
Entrepreneurship offered a different answer. Not instead of social impact - but as a more durable vehicle for it. The organizations that started catching my attention weren't just building strong businesses. They were solving real problems with self-sustaining revenue models and the ability to grow without returning to a foundation with a renewal application every two years. Climate tech. Femtech. Health tech. Govtech. Edtech. Proptech. These are sectors where the business model and the mission are the same thing - where scale isn't just a financial outcome, it's the point. More customers means more impact. A reliable recurring revenue stream means the program doesn't disappear when a funding priority shifts.
Where AI Fits In My World
In tech, I've led customer success, sales, and product -- always focused on how what we do internally ultimately delivers the outcomes and value we want to see with our customers. A lot of my time goes toward improving operations and efficiencies in service of those external outcomes. The internal work only matters because of what it unlocks on the outside.
In the age of AI, that question just got more interesting. What can we automate to create more space for high-value customer work? What vibecoded tools give operators more leverage? Where does AI close the gap between what a team intends to deliver and what actually lands?
The AI Opportunity
Vibecoding - the practice of building functional software tools through plain-language prompts and iteration, no engineering background required - has genuinely changed what it means to build something. I've used it to ship operational tools I couldn't have built otherwise. Real dashboards, real frameworks, real products. The barrier between having an idea and having a working tool has never been lower in my lifetime.
As I found myself reflecting on the potential of vibecoding beyond my day-to-day work, a question started forming: had anyone put these tools in the hands of the problems and spaces I used to think about - the informal workers, the social innovators, the community builders operating on the margins of formal systems? I went looking. And what I found brought everything full circle.
Gringgo, a social enterprise in Indonesia, built an image recognition tool to help informal waste collectors - people who built entire livelihoods sorting debris from waterways and streets - increase recycling rates, better identify what materials are worth selling, and integrate with formal city sanitation systems. The waste pickers I studied in development economics classes, the ones who created self-sustaining micro-economies from what others discarded, now have a tool that meets them where they are and helps them formalize, scale, and grow. The ingenuity was always there. The technology finally caught up.
That wasn't the only example. Better.sg, a Singapore nonprofit dedicated entirely to tech for good, recently used Lovable to build a full production-ready CRM platform for a social enterprise supporting disadvantaged workers entering the food and beverage industry. Eight volunteer developers. Six weeks. A system that can support up to 50,000 users for roughly $70 a month. No engineering budget. No full-time technical staff. Just domain expertise, disciplined prompting, and the right tools.
SuperHumanRace built an AI-powered maternal health app in India that provides rural doctors with real-time, personalized recommendations for high-risk pregnancies -- accurately identifying over 95% of cases in pilots. The founders understood the problem deeply because they were already working inside it. The technology gave them the means to scale what they knew.
These aren't edge cases. They're early signals of something real.
So what happens when you put that capability more broadly in the hands of people who have the most to build and the fewest resources to build it - not as a charity exercise, but as an acceleration layer for people who are already solving real problems?
The social innovator who has a program that works but no capacity to scale it. The community organizer running an initiative out of a group text and a shared folder. The entrepreneur in an emerging market who understands a problem better than anyone because they've lived inside it for years. These are not people who lack ideas or drive. Until very recently, they just lacked the resources and tools.
That gap is closing. And that's the part that genuinely excites me. The ingenuity was never the problem. AI is just the first technology that finally meets it where it lives.
