AI forecasting improves, but superforecasters stay ahead

Human superforecasters continue to outperform artificial intelligence in predicting future events, though the performance gap is narrowing with each competitive quarter. This trend highlights the ongoing evolution of AI capabilities in forecasting—a field with profound implications for financial markets, policy planning, and strategic decision-making across industries.

The competitive landscape: Quarterly tournaments run by prediction website Metaculus show human superforecasters still beat AI systems at forecasting future events through the first quarter of 2025.

• The margin between human and machine performance has consistently narrowed across three consecutive quarters (Q3 2024, Q4 2024, and Q1 2025).
• Metaculus CEO Deger Turan confirms the shrinking performance gap between human experts and AI systems.

Surprising development: The most effective AI forecasting system in Q1 2025 utilized a remarkably simple approach rather than complex methodologies.

• The top-performing model simply retrieved recent news articles and prompted OpenAI’s most advanced model at the time (o1) to make predictions.
• This straightforward approach outperformed many more sophisticated AI forecasting systems, though it still couldn’t match human performance.

Why this matters: The trend suggests AI forecasting capabilities are improving rapidly while human forecasting abilities remain relatively static.

• Forecasting expertise has significant value across domains including financial investments, housing markets, and policy planning.
• As AI forecasting systems continue to evolve, they may eventually match or surpass human capabilities, potentially democratizing access to high-quality predictions.

The human advantage: Superforecasters represent a rare group of individuals with exceptional prediction abilities.

• These experts still have limitations—even superforecasters can be surprised by major events.
• Their forecasting work requires substantial time and effort, and their methods are difficult for average people to replicate.