What will the toolkit of a professional forecaster look like in the 2030s? The recent Metaculus Cup competition, where an AI achieved a top-ten rank, has given us a clear glimpse. The future analyst’s most powerful tool will not be a spreadsheet or a statistical model, but an AI collaborator.
The performance of ManticAI, the British system that placed eighth, showcases the core features of this future tool. First, it will be a multi-agent system, capable of assigning different analytical tasks to specialized AIs. The human user will act as the project manager, directing this team of digital analysts.
Second, it will be a tireless research engine. The AI will handle the 24/7 monitoring of global events, freeing the human expert from the deluge of information and allowing them to focus on strategy and interpretation. It will deliver synthesized briefings and alerts, not just raw data.
Third, it will be a built-in “red teamer” or devil’s advocate. As ManticAI demonstrated with its tendency to disagree with the human consensus, a key function of the AI tool will be to challenge assumptions and prevent groupthink, forcing a more robust and well-rounded analysis.
Finally, it will provide well-calibrated probabilistic outputs, giving the human forecaster a clear, quantitative understanding of uncertainty. The human will then add the final layer of judgment and contextual nuance, especially in situations where the AI’s logic may be flawed. The toolkit of the 2030s is about partnership, combining the best of human and machine intelligence into a single, seamless workflow.