AI’s Judgment Day: Can Machines Develop the “Intuition” to Predict Our Future?

A milestone achievement by a British AI in a forecasting competition is raising a profound question: can machines develop something akin to human intuition? ManticAI’s eighth-place finish in the Metaculus Cup, where it beat many human experts, suggests that the skills once thought to be uniquely human are now being replicated and refined by AI.
The competition required participants to predict the likelihood of 60 future events. The success of ManticAI, which was co-founded by a former Google DeepMind researcher, showcases the incredible progress in AI’s ability to reason, analyze, and synthesize information to make informed judgments about the unknown.
Human forecasters have long relied on a blend of data analysis and a “gut feeling” or intuition. However, professional forecaster Ben Shindel, who competed in the cup, is among those who believe this quality may not be exclusively human. “Intuition is very important, but I don’t think it’s innately human,” he said, suggesting that as AI models become more complex, they may develop emergent abilities that resemble it.
ManticAI’s system operates by deploying a variety of AI agents to assess situations from different angles, gaming out scenarios and researching historical precedents. This persistent, multi-faceted analysis mimics, and in some ways surpasses, the cognitive processes of a human team. Its ability to work on dozens of complex problems at once and constantly learn from new data is a key advantage.
While top forecasters argue that elite humans still outperform bots on average, especially on tasks requiring deep, contextual understanding, the lines are blurring. The rapid ascent of AI in this field suggests that what we call intuition might be, in part, a form of high-speed, complex pattern recognition that machines are becoming increasingly adept at. The future of forecasting will likely depend on harnessing both human and these newly emerging machine “intuitions.”