Definition
Agentic AI refers to artificial intelligence systems that can act with autonomy, making decisions and initiating actions toward a goal rather than only responding to direct instructions. These systems are goal-driven and capable of adapting plans based on new information.
Purpose
The purpose of Agentic AI is to handle tasks requiring initiative and long-term reasoning. It is used in areas where dynamic environments and continuous adaptation are needed, such as workflow automation, research assistance, or robotics.
Importance
- Expands AI beyond reactive systems to proactive decision-making.
- Raises safety and governance concerns due to autonomous actions.
- Useful in environments where constant human supervision is impractical.
- Related to fields like autonomous robotics and multi-agent systems.
How It Works
- Define goals or high-level objectives for the AI system.
- Break down goals into smaller subtasks using reasoning or planning.
- Store and retrieve memory of progress and outcomes.
- Adapt plans based on environmental feedback or new data.
- Execute actions autonomously while respecting constraints.
Examples (Real World)
- AutoGPT: open-source experimental system chaining tasks using large language models.
- LangChain Agents: LLM frameworks that autonomously call tools and APIs.
- Google DeepMind research: exploring autonomous multi-agent systems for problem solving.
References / Further Reading
- Autonomous Intelligent Agents — Stanford Encyclopedia of Philosophy.
- Russell, S., & Norvig, P. Artificial Intelligence: A Modern Approach. Pearson.
- Levels of Autonomy in AI — OECD.AI.
