Kalshi Ramps Up Effort to Build Markets for AI Computing Power

Summary

Kalshi Chief Risk Officer Udesh Jha discusses the launch of a forward curve for AI computing power, aiming to create a benchmark for GPU compute pricing. The product leverages prediction market mechanics for price discovery in a non-standardized, rapidly growing market. Jha highlights hyperscaler spending, expanding AI demand, and Kalshi's ambitions to build a derivatives ecosystem on top of the curve, while also addressing competition with established exchanges and regulatory hurdles.

• Kalshi introduces a forward curve for AI computing power (GPU compute) to establish a market-driven benchmark.
• Udesh Jha argues prediction markets offer superior price discovery compared to traditional futures for nascent, non-standardized assets.
• Hyperscalers are committing $500-600 billion in 2026 capex, signaling enormous future demand for compute.
• Potential users include hyperscalers, neo clouds, AI startups, drug researchers, and retail speculators.
• Kalshi plans to eventually offer swaps, futures, and other derivatives based on the forward curve.
• The company emphasizes its risk-first approach and regulatory compliance, having waited years for approval.
• Jha notes competition from CME and ICE but claims prediction markets provide an edge in accuracy and contract variety.
• Beyond compute, Kalshi is expanding into weather, CPI, crypto, and other risk-management markets to reduce reliance on event contracts.