Imagine a research intern who never sleeps, can read millions of pages in seconds, and can track dozens of complex topics simultaneously. This is the new reality that artificial intelligence is bringing to the world of forecasting and knowledge work, as demonstrated by ManticAI’s recent success in a major competition.
The British startup’s AI, which placed eighth in the Metaculus Cup, essentially acts as the ultimate research assistant. A key to its success is its ability to “work hard persistently,” a trait crucial to effective forecasting. It performs the laborious, time-consuming tasks of data gathering and monitoring at a superhuman scale.
ManticAI’s system deploys a variety of AI agents to fulfill these roles. One agent’s entire job might be to conduct historical research, combing through archives to find precedents for a current event. Another agent is a tireless news-watcher, assessing what is happening now and how it impacts existing forecasts. This frees up the “senior partner”—in this case, the final synthesizing algorithm—to focus on higher-level analysis.
For human professionals, this model is the future. Experts will increasingly be paired with AI assistants that handle the foundational research and analysis. This allows the human to focus on their unique strengths: setting the strategic direction, asking creative questions, interpreting the AI’s findings, and making the final judgment call.
While ManticAI’s system is fully automated, it provides a powerful template for human-AI collaboration. The era of the lone expert spending weeks in a library is being replaced by an era of human strategists guiding their tireless AI interns to uncover insights at an unprecedented speed and scale.
