Computer use lets an AI system work with applications through their visible interfaces. The system can inspect the current screen, choose a relevant control, perform an action, and observe the updated state before deciding what to do next.
This creates a feedback-rich workflow because the agent can inspect and test what it creates. Reliability still depends on accurate perception, current application state, permission boundaries, and whether the agent recognizes and corrects mistakes.
ELI5
Computer use lets an AI agent work through visible software controls much like a person using a screen, mouse, and keyboard. It is useful when an application does not provide a suitable direct tool or programming interface.
For example, an agent can open a settings page and select an approved option. Before clicking, it should refresh the visible state and confirm the correct window because layouts can change and a mistaken action may be difficult to undo.
