glossary

What is agentic AI?

Agentic AI refers to AI systems that act with autonomy — setting sub-goals, using tools, and taking multi-step actions to complete tasks, rather than only responding to prompts.

Agentic AI is the shift from AI that answers to AI that does. An agentic system can break a goal into steps, choose and use the right tools, recover from errors, and keep working toward an outcome with minimal hand-holding.

The term covers everything from coding agents to personal assistants. What makes a system agentic is the loop of planning and acting in the real world — through APIs, browsers, and connected apps — not the model alone.

how this works in folk

Folk applies agentic AI to your personal life: tell it a goal in a text and it plans, uses its own computer and your connected apps, and reports back when it's done.

vanliga frågor

Is agentic AI the same as an AI agent?

They're closely related. 'AI agent' usually means a specific system; 'agentic AI' describes the broader capability of acting autonomously. An AI agent is an instance of agentic AI.

What can agentic AI do today?

It can browse and research, write and run code, manage email and calendars, book services, and run scheduled workflows — like folk booking a restaurant or tracking a flight on its own.

related terms