Vitalik Buterin says he made $70,000 on Polymarket last year by switching into what he called “anti-insanity mode” and betting against implausible outcomes when sentiment overheated. The disclosure came in an interview with Foresight News in Chiang Mai, where he argued crypto’s technology keeps improving while applications drift from the original mission. He framed the profit as a symptom of mispriced narratives, not a personal flex. As prediction markets expand, his remarks are being read as a playbook and a warning for market designers and product risk teams.
— Joe Zhou (@joezhoublack) January 28, 2026
Buterin’s anti-insanity mode and the incentives behind “crazy markets”
Buterin said his edge was spotting markets caught in “crazy mode” and wagering that unlikely events would not materialize. He pointed to bets against Donald Trump winning the Nobel Peace Prize and against the dollar collapsing to zero during panic periods. The approach is governance by skepticism, monetizing the gap between hype and probability. The adoption metrics in the report explain why gaps can widen: Polymarket app installs jumped 1,200% from January to December 2025, and weekly trading volume across platforms climbed from $500 million to nearly $6 billion.
Beyond the trading anecdote, Buterin framed three urgent missions. He said his biggest fear is an industry that devolves into “100% cryptocurrency speculation,” then pushed for more efficient and decentralized Layer 2 networks and a Web2-grade user experience. He also warned that if crypto fails, a future dominated by centralized AI could be “very dangerous.” He cited gas capacity rising from 30 million to 60 million, targeting 300 million, while a MigaLabs report said missed-slot rates climbed above 1.7% when blob counts exceed 16 after December’s Fusaka upgrade.
Oracle risk becomes the limiting factor for prediction markets
He argued that prediction markets were the only sector showing explosive growth in 2025, yet he criticized Polymarket’s emphasis on short term bets such as hourly Bitcoin price moves and weekly sports outcomes. He said these contracts have limited social significance, and incentives should reward longer-horizon forecasting. As an alternative, he referenced Robin Hanson’s Futarchy concept, being tested by MetaDAO, where markets help determine governance methods rather than just settle wagers. Asked about applications, he highlighted decentralized social networking first, then “smarter” DAOs that move beyond simple token voting.
Buterin argued bottleneck is oracle integrity. He recalled a conflict market where Institute for War Studies maps determined control of a train station; a map update made a 5% probability outcome look certain before ISW reversed it next day. He said centralized oracles force trust in one entity, while token-voting systems like UMA can be steered by whales. Still, Polymarket returned to the U.S. after CFTC approval with 10 basis points taker fees and secured $2 billion from Intercontinental Exchange at a nearly $9 billion valuation as adoption accelerated.