On March 20, 2017, ahead of the Galaxy S8 launch, Samsung’s Injong Rhee, Executive Vice President and head of R&D for Software and Services in the Mobile Communications Business, introduced Bixby in a Samsung Newsroom post titled “Bixby: A New Way to Interact with Your Phone.” Rhee framed Bixby around a “conceptually new philosophy”: instead of people learning how a machine works, “it is the machine that needs to learn and adapt to us,” with deep learning built into the core of the interface.
The post argued Bixby was “fundamentally different from other voice agents” because of three properties. Completeness meant that once an application was Bixby-enabled, the assistant could do almost everything the app’s touch interface could do, not just a handful of commands. Context awareness meant a user could invoke Bixby at any point and have it understand the current state of the app and continue the work in progress, mixing voice and touch freely. Cognitive tolerance meant Bixby would accept incomplete or imprecise commands, do what it could, and then prompt for the missing details rather than demanding an exact phrasing.
Samsung also described a dedicated hardware Bixby button on the side of the device to remove the friction of activating the assistant, and said Bixby, running in the cloud, would eventually extend from phones to TVs, appliances, and other connected devices, becoming “an interface for your life.”
For a business reader, Bixby is notable for its bet that the winning assistant would be the one that could drive every function of a device, an ambition aimed at the long-standing problem that people use only a tiny fraction of their gadgets’ features.