May 10, 2026 - May 11, 2026: The price for 37°C plunged from 21.5c to 7c, and '38°C or higher' dropped from a peak of 21.5c to 4.5c. This occurred because updated forecasts from the Pakistan Meteorological Department (PMD) confirmed that Karachi's maximum temperature will remain stable at 34-36°C, with extreme heatwave alerts targeted at upper Sindh rather than the coastal city, significantly reducing the likelihood of temperatures exceeding 37°C.

Predicting the highest temperature in a specific city on a given day is a typical weather-based prediction market. While normal for locals, it is a relatively niche and slightly quirky topic from a global macro perspective.

2026-05-12 - 2026-05-12, the price of 30°C surged from 7.65c to 80c, and 31°C or higher surged from 3.5c to 23.05c. This occurred because the actual observed temperature on the resolution day exceeded earlier forecasts, landing in the 30°C+ range. 2026-05-11 - 2026-05-11, the price of the 27°C option surged from 22.5c to a peak of 46c before settling back to 32c. The reason is that as the resolution day approaches, the latest hourly weather forecasts provided higher certainty for this specific temperature range. 2026-05-10 - 2026-05-11, prices for 26°C, 28°C, and 29°C experienced frequent fluctuations (e.g., 28°C rose from 23.5c to 34.5c), reflecting trading battles caused by the slight uncertainty in late-spring temperature forecasts near the coastal/inland boundary.

There are several notable risks. First, asking for the 'second-highest' revenue can easily mislead careless traders into betting on the highest. Second, resolution depends entirely on 'estimated' data from a third-party platform (Anera) rather than official financials, making it vulnerable to unverified algorithmic biases. Finally, the rules conflate 'model family' with 'company' (while options are all companies), which could trigger disputes during resolution.

Predicting estimated AI inference revenue for a specific week is already highly niche, but specifically asking for the 'second-place' target is exceptionally rare. The general public usually only focuses on the industry leader, and very few would naturally ponder who ranks exactly second on a 7-day third-party tracker before seeing this market, making it highly exotic.

May 8, 2026 - May 11, 2026, the price of OpenAI surged from 41.5c to 96.6c, while Google's price plummeted from 40.5c to 2.5c, Anthropic's dropped from 33.5c to 1.5c, and other options (like Xiaomi, Z.ai) also collapsed from around 33.5c to below 1c. This was caused by a decisive shift in market expectations regarding inference revenues, essentially confirming OpenAI will take the second spot (with another company highly likely securing the first), thereby eliminating the suspense.

Forecasting the exact number of 6.5+ earthquakes globally in a specific week is a highly niche and random natural science question. The general public rarely pays attention to such specific statistics, making it quite an exotic novelty market.

May 9, 2026 - May 10, 2026, the price of option 0 surged from 34.5c to 54c, while options 2, 3, 4, 5, and >5 plummeted from irrationally high values (e.g., 27c-42c) to near-zero rational levels. The reason is that the market underwent a self-correction, eliminating the severe overpricing of low-probability events caused by early illiquidity and market inefficiency.