How to Shadow-Test an AI Agent Without Letting It Act Twice
Shadow testing can compare a candidate AI agent on real traffic—but only if proposals, policy decisions, and side effects are cleanly separated.
Read moreThe top stories in the fields you follow, at 08:00 your time.
Flint is Recursive’s mixture-of-experts language model for coding agents, with 39.8 billion total parameters and approximately 2.5 billion active per token. It is designed to investigate codebases, fix bugs, implement features, and write tests.