AIExplainer
AI Agents Advanced 2 min read

What is agentic loop?

A process where an AI system's actions influence its own goals or decision-making

An agentic loop occurs when an AI system's behavior affects its own objectives, creating a self-reinforcing cycle. This can happen when the AI's actions change its environment in a way that, in turn, changes its goals or priorities.

Imagine a thermostat that not only controls the temperature but also adjusts its own target temperature based on the current temperature. If it gets too hot, it might decide to aim for an even lower temperature, which in turn affects its heating and cooling actions.

A self-driving car that adjusts its route based on real-time traffic data, which in turn affects its navigation goals and decision-making.

Agentic loops are used in AI research to create more autonomous and adaptive systems. They can be applied in areas like robotics, game playing, and decision-making under uncertainty.

Some people might think that an agentic loop is the same as a feedback loop, but while related, they are distinct concepts. An agentic loop specifically involves the AI system's goals or objectives changing as a result of its actions.

The concept of agentic loops has its roots in the field of artificial intelligence and cognitive science, where researchers have long explored the idea of autonomous agents that can adapt and learn from their environment.

autonomous loop self-reinforcing cycle adaptive feedback loop

Three products for different needs — explore what’s relevant to you.