RL Environments Explained: How AI Agents Learn Real-World Work | Brendan Foody, Mercor
Source
youtube.com
Author
Sequoia Capital
Date
Why it matters
Explains why training data for agents is shifting toward RL environments, which shapes how labs improve AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → capabilities and what evalA repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.Full definition → setups look like.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
videoStarcloud's Philip Johnston: Why the Cheapest Compute Will Be in Space
videoBuilding the Automated AGI Lab: Core Automation's Jerry Tworek and Rohan Anil
videoAnthropic's Katelyn Lesse & Angela Jiang: Building an Ecosystem, not a Walled Garden
videoContinual Learning: How AI Agents Get Better With Every Use | Arjun Karanam, Trajectory
videoWhy experts writing AI evals is creating the fastest-growing companies in history | Brendan FoodyLenny's Podcast
videoData and Environment Curation for Post-Training LLMs — Mahesh Sathiamoorthy, Bespoke LabsAI Engineer
videoLearning on the Job: The Future of Post-Training — Raymond Feng, Applied ComputeAI Engineer