-
Book Overview & Buying
-
Table Of Contents
Python Automation Cookbook - Third Edition
By :
As we learned in the previous chapter, giving an AI model access to tools greatly amplifies what it can do. When you combine a reasoning-capable model with the ability to gather new information and carry out actions, you get what's known as a ReAct agent. This is an AI Agent that can Reason about a problem and then Act on its conclusions by using the tools available, hence the name "ReAct." With each result it obtains, the Model can keep reasoning, expanding its capabilities to handle more complex and extended tasks.
These tools, in addition to collecting information, can also execute commands. Command execution can be remarkably powerful, but it's not without risk. The level of access a tool provides determines both the potential and the danger. Greater access can make an agent very effective, but it also increases the chance of unforeseen consequences. AI models can behave in surprising or risky ways.
For instance, imagine...