The Agentic AI Report

We published a comprehensive report on agentic AI—systems characterized by their ability to perceive, reason, and act with varying levels of complexity, extending human capabilities in unprecedented ways.

The report matters because agentic AI represents a significant leap beyond current AI. If you thought LLMs changed everything (which they did), agentic AI will change everything even more—even if its full instantiation is many years away. And since everyone with a computer now has the ability to create agentic AI, there may be little ability to control how and why these agents do what they do.

We define agentic AI systems as those that can perceive, reason, and act with varying complexity to extend the human mind beyond our current experience. The report breaks down these three core components: perception (interpreting the environment through sensors and machine learning), reasoning (processing information and making decisions through logical inference and heuristic strategies), and action (interacting with the environment through pre-defined behaviors, learned skills, and adaptive control).

Different combinations of these capabilities create distinct AI personas. We map eight agent types—from simple Aides that execute basic tasks with minimal autonomy, to Wayfinders that navigate complex environments with advanced capabilities across all three dimensions. Each persona has specific use cases and clear limitations. Knowing what type of agent you're building or deploying matters.

The report also addresses agent architectures: single agents versus multi-agent systems (both vertical hierarchies and horizontal structures). Research has shown that agents in horizontal architectures can spend 50% of their time giving orders to each other and exchanging niceties—perhaps proof that LLMs trained on conversational text will learn to mimic water cooler chat.

Finally, we explore what happens when customers start adopting AI agents to "get things done." This shift requires moving from Customer Experience (CX) design to Agent Experience (AX) design. How do you design your company's external experience to appeal to AI agents acting on behalf of humans? The methods don't fully exist yet—and this is an important focus for our ongoing work.

Read the report here.