From Digital Assistants to Digital Companions Featured Pattern: P0869 January 2016

Digital assistants might travel among applications and devices to follow users regardless of where they are or what they are doing.

Abstracts in this Pattern: (Seattle, Washington), Apple (Cupertino, California), Facebook (Menlo Park, California), Google (Mountain View, California), and Microsoft (Redmond, Washington) each has its own proprietary digital assistant that locks users into its own ecosystem. Missing features and a lack of expandability increase the likelihood that users will have complex interactions with multiple disparate digital assistants. For instance, 11 European research groups (Semeoticons; are working on the Wize Mirror—a smart mirror that can read a number of biometric factors and may be able to diagnose some diseases via a multispectral camera and chemical-detection technologies. Researchers hope the Wize Mirror will help users remember to take their medicine and monitor their health. The device may even act as an assistant to people in aging-in-place arrangements.

Proprietary digital assistants will likely stifle innovation to lock users into a walled garden of devices and services under the control of a single company. Open platforms for accessing, storing, and sharing data among digital-assistant services may enable far greater innovation and give start-ups the opportunity to thrive. Email has remained ubiquitous, open, adaptable, and democratic, and it holds data persistently; therefore, digital assistants could source their data from emails. For example, start-up Clara Labs (San Francisco, California) has developed Clara—a platform-agnostic scheduling system that allows users to schedule appointments by cc'ing Clara (the artificial intelligence) on their emails. Clara also peeks into users' calendars and sends emails to schedule meetings. Currently at the beta stage, Clara already can integrate with any email client and any online calendar system.

In the future, digital assistants might work much more closely with users than they do now, becoming very personal companions that individuals can program according to their specific needs. Researchers at University College London (London, England) are working to create a natural-language programming language that would enable people to use natural language rather than code to change an application's characteristics.