From a Hacker News comment, as a response to someone’s hand-wringing, comparing all of the new AI tools, features, and conceptual notions to the Javascript framework madness pre-React. The key insight is:

The foundational mental model to get the hang of is really just:

• An LLM
• …called in a loop
• …maintaining a history of stuff it’s done in the session (the “context”)
• …with access to tool calls to do things. Like, read files, write files, call bash, etc.

Some people call this “the agentic loop.”

Seems simple in an obvious sort of way, but it very succinctly summarizes how these pieces fit together, while also suggesting the key superpower to fully unlocking the power of LLMs - namely, wrapping it all in a loop.