Toby Shevlane: Teaching AI to Think Beyond the Crowd
Written by Admin • 2026-09-21 00:46:00
Toby Shevlane and Mantic are exploring a new frontier of artificial intelligence, focusing on systems designed to handle uncertainty and improve forecasting. Their recent success in the 2026 Metaculus Cup highlights the potential of predictive AI in various sectors.
Toby Shevlane: Teaching AI to Think Beyond the Crowd
Exploring a new frontier in artificial intelligence with Mantic
From Generative AI to Predictive AI
As artificial intelligence continues to evolve, the focus is shifting from generative capabilities—such as writing and image creation—to predictive abilities. Toby Shevlane, co-founder of Mantic, aims to redefine how AI handles uncertainty. Founded in 2024, Mantic is not just another AI company; it is pioneering a new class of systems that leverage historical data and advanced algorithms to make forecasts that can rival human judgment. This transition represents a fundamental shift in AI's role, moving from merely generating outputs to intelligently navigating the complexities of future possibilities.
Mantic's approach challenges traditional AI narratives by emphasizing the importance of decision-making under uncertainty. The company’s recent success in the 2026 Metaculus Cup, where its AI outperformed human forecasters, underscores the potential of predictive AI. As organizations grapple with increasingly complex data and rapid change, the ability to anticipate outcomes and evaluate probabilities becomes invaluable, marking a significant step toward smarter AI applications in various sectors.
Why Forecasting Is So Difficult
Forecasting is inherently challenging due to the unpredictable nature of many variables that influence outcomes. Historical data can provide insights, but the ability to account for new, unforeseen factors is what distinguishes effective predictive systems. Mantic aims to tackle this complexity using cutting-edge AI technology. By refining models with iterative learning and feedback, the company enhances the reliability of its forecasts, allowing organizations to make more informed decisions when faced with uncertainty.
The difficulty of forecasting is often compounded by the human tendency to rely on consensus and the comfort of established norms. Mantic's technology seeks to bridge this gap by offering a tool that not only predicts but also challenges conventional wisdom. This capability can significantly impact sectors such as finance, policy-making, and strategic planning, where precise forecasting can lead to better resource allocation and risk management.
Inside Mantic's Approach
Mantic's approach to AI forecasting combines the latest in machine learning with real-world applications. By integrating frontier AI models with extensive historical datasets, the company enhances the predictive power of its systems. This method allows Mantic not only to generate forecasts but also to refine them continuously based on emerging data and outcomes, effectively learning from past predictions. The iterative nature of this process sets Mantic apart in the crowded AI landscape.
The technology developed by Mantic has caught the attention of various industries, including finance and government. Its potential to improve decision-making processes by providing a clearer picture of uncertainty positions Mantic as a leader in the predictive AI field. This capability is not just about making predictions; it's about transforming how organizations understand and respond to the unpredictable nature of the future.
The next frontier of AI may not be generating answers, but navigating uncertainty.— Toby Shevlane
Facts & Insights
- Mantic was co-founded in 2024 by Toby Shevlane, a former Google DeepMind researcher, and Ben Day.
- The company raised US$25 million in seed funding in September 2026 in a round led by Radical Ventures.
- Mantic's AI performed strongly in the 2026 Metaculus Cup, a forecasting competition involving human participants.
The Intelligence of Uncertainty
The most valuable AI systems may not always be the ones that produce the fastest answer. They may be the ones that help people understand how uncertain the answer actually is. Toby Shevlane's work with Mantic reflects this emerging direction, where artificial intelligence becomes a tool for evaluating possibilities rather than simply generating content. As businesses and institutions face increasingly complex decisions, the ability to reason systematically about uncertainty could become an important new layer of digital intelligence.
In conclusion, the pursuit of improved forecasting capabilities represents a significant advancement in AI technology. Mantic's innovative approach not only addresses the challenges of uncertainty but also paves the way for more effective decision-making across various sectors. As organizations continue to navigate an unpredictable world, the insights provided by predictive AI could prove to be invaluable.
$25M
Seed Funding
2024
Founded
2
Co-Founders
1
Core Mission: Forecasting
Categories
Tagged in this story
Image sources
Pexels - Photo by Google DeepMind