Scott analyzes whether AI forecasters will plateau near human levels or keep improving, arguing that even non-miraculous progress could yield 4-12 percentage point improvements in prediction market accuracy. Longer summary

Scott examines whether AI forecasters will plateau near current human performance or continue improving far beyond it. He analyzes a 2010 study by Daniel Reeves showing prediction markets only marginally outperformed simple statistical models, but argues this doesn't necessarily limit future progress. The post reframes those results to show they actually represent meaningful gains in difficult-to-predict domains like sports, then applies this reasoning to geopolitical forecasting. Using three different 'anchors' - comparing AI improvements to past technological gains, human-to-superhuman forecaster improvements, and chess AI progress - Scott estimates that maximally capable AI forecasters could improve prediction market accuracy by 4-12 percentage points over current levels, which he argues would be genuinely valuable even if not miraculous. Shorter summary

Scott Alexander discusses recent developments in prediction markets, including midterm forecasts, legal challenges, nuclear risk assessments, and new market applications. Longer summary

This post covers various topics related to prediction markets and forecasting, including midterm election predictions, the CFTC's actions against PredictIt, nuclear risk forecasts, Kalshi's application for election markets, and updates on various prediction markets and forecasts. Scott discusses the discrepancies between poll-based and prediction market-based forecasts for the US midterms, the legal challenges to the CFTC's decision to shut down PredictIt, and recent nuclear risk assessments by forecasting groups. He also covers Kalshi's efforts to gain approval for election markets and provides updates on several ongoing prediction markets. Shorter summary