Forecast LabsThere Is an Optimal Path. We Find It.
We deploy elite forecasting agents with predictive power on par with top human superforecasters, designed to discover and chart the optimal course for humanity to harness the benefits of AI and reduce catastrophic risks.
AI Forecasters Outperform on Prediction Markets
DelPy demonstrated significantly higher accuracy than human experts on Manifold Markets, with a 38% improvement in prediction accuracy. This represents a major step forward in AI-assisted decision making for AI Safety.
From Forecasting to Action
Our advanced AI systems use causal models to understand complex relationships and make accurate predictions about critical events. By focusing on AI Safety, we're working to reduce existential risks and ensure AI development benefits all of humanity.
Every prediction we make is a step toward better decision-making in high-stakes scenarios, from AI governance to global risk assessment. Together, we can build a safer future.
A causal model explaining the veto of California Senate Bill 1047. The bill, which aimed to regulate the development of large-scale AI models, ultimately failed due to significant concerns that it posed a major threat to open-source AI, would stifle innovation, and was built on a flawed and burdensome regulatory framework. Note probabilities here are not actual probabilities predicted by DelPy, they are here for reference.
Ready to Explore AI Safety Forecasting?
Connect with our team to learn more about DelPy, our AI forecasting system, or discuss collaboration opportunities in AI Safety research.