Start‑ups Harness AI to Peer Into Tomorrow’s Markets and Crises

It feels a bit like science‑fiction, but the reality is that a handful of start‑ups are already feeding massive data streams into clever algorithms and, in many cases, nudging human experts off the pedestal they once occupied. The goal? Not to claim certainty – no one can promise the future – but to deliver a probability that’s sharper than anything a person could conjure in a boardroom.

Take the Metaculus Cup, for instance. In early September two AI models swooped in and seized the top two spots in a competition that pits prediction champions – both flesh‑and‑blood and silicon‑based – against each other. It was a headline‑grabber for a reason: these systems were matching, and sometimes beating, the world’s best “super‑forecasters.”

Dan Schwarz, co‑founder of the forecasting firm FutureSearch, says the shift will feel like stepping out of the Stone Age. “By 2030 we’ll look back and wonder how we ever made decisions without AI‑driven odds,” he told reporters. The company’s in‑house wizard, Tom Liptay – a recognised super‑forecaster – admits he feels a twinge of anxiety. “I know my job isn’t safe in the medium term,” Schwarz chuckles, “because the models are learning to out‑think even the best of us.”

What makes these AI models tick is their appetite for data. Public statistics, private market feeds, satellite imagery, even social‑media chatter – everything is gobbled up, cleaned and turned into patterns that humans would miss. The result is a probability curve that can be layered with other events. Want to know how a sudden spike in oil prices might affect a conflict in the Middle East? The AI can simulate the ripple, offering a quantified outlook rather than a vague gut feeling.

It isn’t just academia or giant hedge funds getting a slice of the pie. Theo Summer, who co‑founded the start‑up Preseen with Veniamin Veselovsky, turned a modest $35 into $1.9 million in just seven months on the US prediction‑market platform Kalshi. Their secret sauce? AI‑driven betting signals that spot mispriced outcomes before the crowd catches on.

Veselovsky notes that “the quickest buyers are usually hedge funds because they can move fast when they see a chance.” Yet even those behemoths are feeling the pressure. Algorithms have been whispering to Wall Street for a decade, but the rise of sophisticated AI forecasting promises to turn whispers into full‑blown conversations, automating decisions that used to require teams of analysts.

Of course, there are caveats. An AI can only be as good as the data it receives, and bias in those inputs can skew forecasts. Start‑ups are therefore betting not just on raw computational power, but on robust data‑governance and transparent model‑testing. The idea is to give investors, policymakers and businesses a clearer lens, not a crystal ball.

As the technology matures, we may soon see AI forecasts embedded in everything from supply‑chain planning to election night coverage. The real gamble isn’t whether the machines can predict the future – it’s whether we’ll learn to trust the probabilities they hand us.