logo

Overview

As the pace of AI accelerates globally, it is becoming clear that Agentic AI will be one of the main forms of developing AI systems. As the name implies, it is a combination of two concepts - Agents and AI. The goal of this website is to collate, ideate and disseminate various aspects of Agentic AI.

Agents

The English word “agent” is derived from the Latin “agens”, simply meaning “one who acts”. According to Collins Dictionary, agent is a noun meaning “a person who acts on behalf of another person, group, business, government, etc; representative”.

The concept of software based gents have been around for a few decades. One of the earliest papers on the topic is Intelligent Agents: Theory and Practice. Another free textbook is Artificial Intelligence: Foundations of Computational Agents, 3rd Edition. Both of these books serve as good starting points on the topic of agents.

A simple definition of an agent could be “A software entity that perceives its environment and acts upon that environment’. It is a fairly broad definition and could be interpreted and implemented in many ways depending on the use cases.

AI

It was in the 1950’s, the term Artificial Intelligence (AI) was coined. John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon proposed the term “artificial intelligence” at a conference in 1955.

Over the next few decades, AI went through periods of hype and disillusionment. Some key developments were -

  1. Logic Theorist (1956) - A computer program written by Allen Newell, Herbert A. Simon, and Cliff Shaw. It was the first program deliberately engineered to perform automated reasoning, and has been described as “the first artificial intelligence” program.
  2. ELIZA (1966) - Joseph Weizenbaum’s chatbot simulated a Rogerian psychotherapist, demonstrating early natural language processing capabilities.
  3. Shakey (1966-1972) - Developed at SRI International, Shakey was one of the first robots capable of navigating and interacting with its environment.
  4. Expert systems (1980s): These systems, designed to mimic the decision-making of human experts in specific domains, gained popularity in the 1980s.
  5. Deep Blue (1997) - IBM’s Deep Blue defeated world chess champion Garry Kasparov, demonstrating the power of AI in complex game playing.
  6. Deep Learning (2000s) - Deep learning, a subfield of machine learning using artificial neural networks with multiple layers, began to show its potential.
  7. AlphaGo (2015) - A breakthrough in artificial intelligence by demonstrating the ability of a machine learning system to master the complex game of Go, achieving a level of play that surpasses even top human players, primarily through its innovative combination of deep neural networks and reinforcement learning, which allows it to learn and improve strategies through trial and error, rather than relying solely on pre-programmed rules

Agentic AI

This brings us to the most nascent advancement being created on top of agents and AI - Agentic AI. It can be though of as “artificial intelligence systems designed to act autonomously within an environment, capable of making decisions and performing actions with minimal human intervention.”