Anthropic's Boris Cherny on loop engineering with AMD
YouTube
Anthropic's Boris Cherny discusses loop engineering, a methodology that shifts from traditional coding to managing AI agents through graph-based workflows that cycle through planning, action, verification, and retry or escalation steps. The approach structures agent behavior around loops rather than linear code execution. This changes how developers design AI systems from writing sequential instructions to orchestrating iterative agent decision cycles.
Why it matters
Loop engineering represents an abstraction shift from writing code to managing agents by designing graph workflows: plan → act → verify → retry or escalate → done.
