Epoch AI is a research organization dedicated to producing rigorous, data-driven forecasts and analysis about artificial intelligence progress, with particular focus on compute trends, training datasets, algorithmic efficiency, and AI timelines.
Facts
1Divisions
2Maintains the Parameter, Compute and Data Trends database — one of the most cited datasets on ML model scaling. Tracks 700+ notable ML systems.
Research on AI trends, forecasting, and compute scaling. Publishes influential analyses on training compute trends, parameter counts, and dataset sizes.
Related Wiki Pages
Top Related Pages
AI Timelines
Forecasts and debates about when transformative AI capabilities will be developed
AI Scaling Laws
Empirical relationships between compute, data, parameters, and AI performance
AI Compute Scaling Metrics
Empirical tracking of measurable indicators of AI compute scaling: training compute growth, GPU deployment, capital expenditure, inference shifts, ...
AI Risk Critical Uncertainties Model
This model identifies 35 high-leverage uncertainties in AI risk across compute, governance, and capabilities domains.
Capability-Alignment Race Model
This model analyzes the critical gap between AI capability progress and safety/governance readiness. Currently, capabilities are ~3 years ahead of ...