AI Agents Are Just Distributed Systems Now — Salman Munaf, TikTok
Source
youtube.com
Author
AI Engineer
Date
Context
An 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 → that times out while requesting a refund cannot safely assume the refund failed. The service may have completed it without returning a response. The example in this reliability talk's published description explains why simply retrying can pay twice. A request identifier lets the system look up what happened; an idempotency key lets repeated requests represent the same operation instead of separate refunds. These controls give the agent a way to resolve uncertainty before repeating a consequential action. Reliability therefore depends on how the surrounding services handle uncertain outcomes, as well as on the model's decisions.
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.
determinism — Whether the same input reliably produces the same output — something LLM systems mostly lack, which changes how you test and debug them.
video1,000 Agent Tasks in a Sandbox: What Breaks When LLMs Run Code — Kevin Orellana, AWS
videoHow Inference Keeps Up with Our Insatiable Need for Tokens — Cerebras & Gimlet Labs
videoAuth for Agents: Let AI Sign Up for Your Product with auth.md — Michael Grinich, WorkOS
videoDecoupling the Brain from the Hands: Claude Managed Agents (Part 1 of 4) — Anthropic
