Computerized task guidance across devices and applications

A computer-implemented method, comprising:

• providing a first computing device and a second computing device access to a working task document including a textual task description describing task steps to be completed in performing a task;
• generating a computer-readable task state data structure that represents a state of interaction with the working task document, with the task state data structure being accessible via the first computing device and the second computing device;
• in response to receiving information describing a visual user input received via the first computing device, performing a first update to the task state data structure to represent the visual user input, with the visual user input indicating a state of progress in completing the task;
• and
• in response to receiving information describing a natural language user input command received via the second computing device, performing a second update to the task state data structure to represent the natural language user input, with the natural language user input indicating a state of progress in completing the task;
• and
• providing the second computing device access to the task state data structure to facilitate generation of a natural language response.

A computer implemented method comprising:

• conducting an interactive computer communication session via a visual interface component, an audible interface component, a shared working task document, and a shared task state data structure, with the working task document comprising instructions for performing a task, with the task state data structure indicating a state of completion of the task, and with the conducting of the session comprising:
• performing a first update of the task state data structure in response to visual user input received from the visual interface component, with the visual user input directed at an element of the working task document;
• performing a second update of the task state data structure in response to natural language input received from the audible interface component;
• providing a first set of task guidance to the visual interface component in a visual format, with the first set of task guidance being provided using the working task document and the task state data structure, and with the first set of task guidance representing the first update and the second update;
• and
• providing a second set of task guidance to the audible interface component in a natural language dialog format, with the second set of task guidance being provided using the working task document and the task state data structure.