From the course: Agentic AI Architecture Foundations: Designing Autonomous AI Systems
Join today to access over 24,800 courses taught by industry experts.
The agent-environment interaction loop
From the course: Agentic AI Architecture Foundations: Designing Autonomous AI Systems
The agent-environment interaction loop
In this video, we'll explain how agents relate to their surroundings. A core part of this relationship is the agent environment interaction loop, which is an essential iterative cycle that agents participate in to understand the world around them and to then alter that world with their actions. Before we get into the description of the loop and its steps, we first need to establish the concept of an environment as it relates to agentic AI. The environment for an agent is essentially the surroundings that it finds itself residing in and interacting with, for example, an environment for an agent that is focused on processing customer requests and orders might consist of customer data, marketing information, e-mail communication, and product inventory. It perceives this environment based on the digital data it receives as input in relation to each customer activity. An environment for an agent that is embedded in a robot, on the other hand, might consist of a room in a warehouse with a…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.