Matthew Kenney is the founder and CEO of the Algorithmic Research Group (ARG), a small research lab in Durham, NC studying how software systems recursively improve themselves. His primary focus is building rigorous benchmarks to measure AI agents' ability to conduct machine learning research tasks, with support from Open Philanthropy. Key outputs include the ML Research Benchmark (MLRB), DeltaMLBench, and the ARIA Benchmark. His work sits at the intersection of AI capabilities measurement and AI safety, aiming to understand the current limits of AI-driven R&D before such systems become transformative.
Funding Details
- Annual Budget
- -
- Monthly Burn Rate
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- -
Theory of Change
By building rigorous benchmarks that accurately measure how capable AI agents are at conducting machine learning research, Kenney aims to give the AI safety community an early warning system for when AI-driven R&D becomes a transformative risk factor. Understanding the current limits of self-improving AI systems — and tracking how those limits change — allows researchers and policymakers to anticipate dangerous capability jumps and respond proactively rather than reactively.
Grants Received
from Open Philanthropy
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 9:51 PM UTC
- Created
- Mar 20, 2026, 2:34 AM UTC