Agentic Software System

From GM-RKB
(Redirected from Software agent)
Jump to navigation Jump to search

An Agentic Software System is an autonomous goal-directed AI-based software system that can support agentic task executions.

References

2024-11-24

[1] https://en.wikipedia.org/wiki/Agent-based_model
[2] https://botpress.com/blog/real-world-applications-of-ai-agents
[3] https://www.mdpi.com/2078-2489/11/5/271
[4] https://builtin.com/articles/agent-based-modeling

2023

  • GBard
    • An agent is a system that can perceive its environment, make decisions and take actions to achieve a specific goal or set of goals. The agent operates autonomously, meaning it is not directly controlled by a human operator.
    • The four rules that all AI agents must adhere to are:
      • Agent Perception: The agent must be able to perceive the environment. This means it must have sensors that can collect information about the environment, such as cameras, microphones, and sensors that measure temperature, humidity, and light.
      • Agent Decision-making: The agent must be able to make decisions based on the information it perceives. This means it must have a decision-making algorithm that can weigh the pros and cons of different actions and choose the one that is most likely to achieve the goal.
      • Agent Action: The agent must be able to take actions to change the environment. This means it must have actuators that can perform actions, such as moving its limbs, driving a car, or controlling a robot.
      • Agent Rational action: The agent's actions must be rational. This means they must be based on the agent's understanding of the environment and its goals.
    • In addition to these four rules, there are other characteristics that are often associated with intelligent agents. These include:
    • The concept of an intelligent agent is a broad one, and there are many different types of agents that have been developed. Some examples of intelligent agents include:
      • Game playing agents: These agents are designed to play games against humans or other agents.
      • Search agents: These agents are designed to find information or objects in a large search space.
      • Planning agents: These agents are designed to plan sequences of actions to achieve a goal.
      • Robotic agents: These agents are designed to control robots and other physical systems.
      • Virtual agents: These agents are designed to interact with humans in a virtual environment.

2022

  • (Sutton, 2022) => Richard S. Sutton. (2022). “The Quest for a Common Model of the Intelligent Decision Maker". In: Proceedings of the Fifth Multi-disciplinary Conference on Reinforcement Learning and Decision Making
    • ABSTRACT: The premise of the Multi-disciplinary Conference on Reinforcement Learning and Decision Making is that multiple disciplines share an interest in goal-directed decision making over time. The idea of this paper is to sharpen and deepen this premise by proposing a perspective on the decision maker that is substantive and widely held across psychology, artificial intelligence, economics, control theory, and neuroscience, which I call the "common model of the intelligent agent". The common model does not include anything specific to any organism, world, or application domain. The common model does include aspects of the decision maker's interaction with its world (there must be input and output, and a goal) and internal components of the decision maker (for perception, decision-making, internal evaluation, and a world model). I identify these aspects and components, note that they are given different names in different disciplines but refer essentially to the same ideas, and discuss the challenges and benefits of devising a neutral terminology that can be used across disciplines. It is time to recognize and build on the convergence of multiple diverse disciplines on a substantive common model of the intelligent agent.
Retrieved from "http://www.gabormelli.com/RKB/index.php?title=Agentic_Software_System&oldid=951017"