#预测市场 Kalshi's prediction market is quite interesting—Gemini's probability has skyrocketed from 30% at the beginning of the year to 86%. The logic behind this shift in probability is worth analyzing. The trading volume of $14.08 million also indicates that market participation is indeed heating up, with many people voting with real money.

My observation is that data changes in such prediction markets often reflect a re-pricing of market expectations. Gemini's jump from 30% to 86% suggests either a significant breakthrough in product capability or that market sentiment is collectively optimistic about a certain direction. The key point is that settlement is based on LM Arena rankings, which means that ultimately, actual performance data will speak for itself—leaving no room for superficial packaging.

For traders who like to follow market hotspots, these prediction markets can serve as a reference signal. Of course, one should not blindly follow the option with the highest probability; instead, focus on those possibilities that are seriously undervalued—this is the true value of prediction markets. The fact that ChatGPT and Grok's probabilities are pushed so low doesn't necessarily mean there's no comeback opportunity; it depends on how you assess the speed of technological iteration.

Prediction markets essentially test one's cognition and ability to weight and interpret information, which is quite similar to trading.