The 2029 Tipping Point: When AI Could Out-Forecast Humanity’s Best

A specific date is now on the horizon for potential AI supremacy in forecasting: 2029. This prediction, made by the CEO of forecasting platform Metaculus, comes in the wake of a stunning performance by a British AI in their annual competition. ManticAI’s eighth-place finish has provided compelling evidence that AI is on a trajectory to match or even surpass the best human forecasters within the decade.
The Metaculus Cup, which concluded this September, saw ManticAI’s system out-predict numerous human experts on 60 challenging real-world questions. This result, a massive leap from the performance of bots in previous years, prompted Metaculus CEO Deger Turan to put a timeline on AI’s ascent. He estimated that by 2029, AI would be “on a par or better” than top human forecasters.
This forecast is based on the exponential rate of improvement in AI capabilities. A year ago, the top AI was ranked around 300th. Today, multiple AIs are in the top tier. If this pace of development continues, AI systems will rapidly close the remaining gap with elite human “superforecasters.”
The AI’s current strengths are in its persistence and data-processing power, using a team of AI agents to work tirelessly on many problems at once. Its primary weakness remains in nuanced, multi-stage logical reasoning, where humans still excel. However, advances in AI architecture are addressing these limitations at a rapid pace.
The 2029 prediction serves as a landmark for the field. It sets a concrete goal and a timeline for the development of highly advanced predictive AI. As we approach this potential tipping point, the focus will increasingly shift from competition to collaboration, as humanity prepares to integrate this powerful new form of intelligence into its decision-making frameworks.