TLDR Tech / Crypto
How AI Will Save Prediction Markets
About 65% of volume on Polymarket and Kalshi concentrates in sports, while STEM markets account for just 1.2%, a distribution that contradicts Robin Hanson's Futarchy thesis and Vitalik Buterin's Info Finance vision of prediction markets as societal truth machines. The four trader archetypes explain the clustering: gamblers favor short-horizon sports bets, sharps follow gamblers per the no-trade theorem, savers avoid zero-sum exposure, and hedgers participate only in narrow verticals, leaving subsidizers trapped in a VOI/MVL tradeoff where high-information markets require outsized liquidity bootstraps relative to expected information value. The Iran strike timing markets demonstrated both the ceiling and the risk of this dynamic, reaching $529M in volume before a special forces soldier was arrested for insider trading on Polymarket, while regulators in Nevada, Arizona, and Utah have moved to ban, criminalize, or formally designate prediction market activity as gambling. AI agents are the structural remedy: always-on, low-cost participants that can act as both market makers and informed traders in niche markets, reducing the MVL threshold enough to make non-sports prediction markets economically viable.
Read the full story