#预测市场 Seeing Coinbase CEO's remarks, what flashed through my mind were the DAO incident of 2016 and a series of subsequent predictions. Back then, we were still debating what blockchain could do; who would have thought that ten years later, prediction markets would become a more effective signal mechanism than traditional polls.

Polymarket's retention data is very interesting—an 85% retention rate crushing DeFi platforms, and that's no luck. I've seen too many projects die from the "user acquisition trap," where subsidies are poured in for three months only to become dead projects. But prediction markets are different; they inherently possess driving force. Every election, every macroeconomic data point is a real trading reason, without relying on airdrops and incentive games to keep users engaged.

This reminds me of the ICO bubble in 2017 and DeFi Summer in 2020. At that time, what we saw was speculative enthusiasm but a lack of sustainable application logic. Prediction markets, on the other hand, are the opposite—they derive their value not from token price speculation but from providing a new way to price information. Policymakers will really use this, and the media will pay attention because there are real incentives and constraints involved.

Coinbase, Phantom, Gemini are all making moves, indicating that the industry has shifted from "proof of concept" to "institutionalization." This cycle is different from ten years ago. We're no longer gambling on the future but reviewing past lessons to find applications that can truly sustain themselves. Prediction markets might be the answer in this round.