AI Prediction Markets Are Broken — And That's Where the Edge Is
Everyone's trading AGI timelines. Almost nobody understands what they're betting on.
The future isn't predicted. It's priced.
There's a war happening in prediction markets right now, and most people don't even know it exists.
On one side, you have Metaculus — the forecasting platform beloved by rationalists and AI researchers — predicting AGI arrival by 2030 with high confidence.
On the other side, you have Polymarket, where actual money is on the line, showing significantly different probabilities.
Someone is catastrophically wrong.
The Information Asymmetry
Here's what makes this fascinating: the Metaculus community is arguably the most informed group of forecasters on the planet when it comes to AI. These are researchers, engineers, and professional forecasters who eat base rates for breakfast.
But Polymarket has something Metaculus doesn't: skin in the game.
And as Nassim Taleb would remind us, the person with skin in the game processes information differently than the person without it. When your money is on the line, you think harder about what you don't know.
Where the Edge Lives
The edge in AI prediction markets isn't in having better information about AI progress. Everyone reads the same papers, follows the same labs, watches the same benchmarks.
The edge is in understanding what the market is actually pricing in versus what participants think they're pricing in.
Most AI prediction market traders are pricing in a narrative — "AI is progressing fast, therefore AGI soon." But narratives aren't probabilities. They're stories. And stories have a nasty habit of being wrong in ways that feel right.
The smart money isn't asking "when will AGI arrive?" It's asking "what does the current price imply about the speed of progress required, and is that speed consistent with historical rates of technological development?"
That's a much harder question. And that's exactly why the answer has edge.
The Reality Check
If you're trading AI prediction markets, here's your homework: define your terms. What exactly does "AGI" mean in the market you're trading? What are the resolution criteria? What's the historical base rate of transformative technology predictions being right?
Most traders can't answer these questions. That's not a criticism — it's an opportunity.
The signal is there. Are you reading it, or just following the crowd?