Prediction markets are going to be very large. I expect them to grow to many times the size of today’s stock market, because what you can trade isn’t limited to the few thousand companies that happen to be publicly listed. You can trade real-world outcomes: tomorrow’s weather, the price of gas, an election, an interest-rate decision.
People have always been willing to put money on outcomes. The lottery is the plainest example, and it’s a market where nobody has an edge. Prediction markets are the version where knowledge counts.
That changes who gets paid for what they know. Soon an expert in almost any field will be able to put a price on their expertise and be rewarded for being right. I’ve long believed you don’t really know a subject until you can predict the likelihood of its outcomes. Prediction markets test exactly that.
I’ve always liked investing because it’s the best form of education I know, with the exams built in. Losing money is ongoing tuition. Making money is the reward for doing it well, and the fuel to sit the next exam.
Prediction markets open that loop to everyone, on whatever they’re curious about. I know nothing about Taylor Swift, but fans who know how good her latest single is can put that knowledge to the test by trading on how popular it will be. Pick a subject you care about, learn it, get tested on it, and get paid when you’re right. The incentive feeds the curiosity, and the curiosity sharpens the next trade.
The tools haven’t caught up. Platforms for trading stocks, bonds and options are sophisticated, built up over decades. Prediction markets are still young. When I started trading them, I found the tools lacking: the apps show you the price and little else. Even mature trading platforms lack depth on the data behind an outcome.
Who I am
I’ve traded stocks and invested for decades. Before building software, I spent years as a Wall Street analyst, forecasting company financials for a living. Today I build new AI products. Pretickted sits where those three meet.
Why now
AI tools are improving quickly, and there is serious work on “world models”, AI that learns how the real world behaves. I think we’re laying the foundation for something new: a constant exchange of information between the expertise in people’s heads and the events happening in the world, with prediction markets as the place where the two meet.
There’s also something here that has never been true of stocks, crypto or any other traditional asset. In a market on a real-world outcome, an ordinary person can have more information than a large institution. Someone who watches gas prices in their town, or follows the weather closely, can know more than a fund does. Pretickted is built to give that person the best data available.
Why we started with gasoline
I sort the information behind any trade into two buckets: what’s knowable and what isn’t. In gasoline, the knowable bucket is unusually full.
- It settles on what regular people actually pay at the pump each day, which makes it hard for any single company or group to manipulate.
- It has a live proxy. GasBuddy measures the same average throughout the day, and there are leading indicators such as wholesale gasoline prices.
- It has patterns. All else equal, gas tends to be cheapest on Sunday evening and most expensive in Thursday rush hour.
Taken together, a trade like gas carries more real-world signal than a bet on what Google’s stock will do today, where flows of institutional money matter far more than anything happening in the world.
What it has done for me
This year I put $250 into a Kalshi account and used Pretickted to inform my trades. As of October 2026, that account has grown about 30 times.
This isn’t a brag, or a promise. It was the validation I needed to see for myself to keep building the site. Before this, I knew nothing about gas. Every losing trade was tuition. I learned what information Pretickted could have gathered and synthesized for me, so I wouldn’t make the same mistake twice if it could be avoided.
What we’re building next
- Your account, on our screens. Connect your Kalshi account to see and manage your actual offers and positions next to the data.
- AI forecasts. We’re introducing AI models that predict a reasonable set of values for the next readings. It’s early. Today an AI’s guess is no better than a person’s, and we treat it as one more data point. But its ability to recognize patterns is improving fast, and I expect it to match people’s soon.
- More markets. Each one gets the same treatment as gas: its real-world source, its full history, and forecasts.
If you trade prediction markets, I’d like you to try it.