AI Agent

A chatbot takes a prompt and returns a response. An agent is given a goal and runs a loop: it plans an approach, calls tools to take real actions (search, run code, hit an API), observes what came back, and decides the next step — repeating until the goal is met or it gives up. That loop, not the model itself, is what makes something “agentic.”

The capability that unlocks agents is the model deciding when and how to act, not a human scripting every branch. That’s also what makes them hard to ship: more autonomy means more ways to go wrong, so production agents lean heavily on guardrails, evaluation, and memory to stay reliable. Much of the show is builders working through exactly that gap between a demo that works once and an agent that works every time.