It’s tempting–and not entirely unreasonable–to lump Apple’s Siri, Google’s Google Now, and Microsoft’s Cortana into one technological bucket and call them voice assistants. But the interesting thing about the three services isn’t what they have in common so much as how they’re diverging.
At Microsoft’s BUILD conference keynote here in San Francisco, one of the demos involved Cortana being able to figure out a user’s plans from discussions in Skype and helpfully offering to do things like order food and reserve a hotel room. It does so through sophisticated exchange of information with third-party bots that are designed to handle such specific tasks as well as possible.
After the keynote, I met up with Marcus Ash, the Microsoft group product manager in charge of Cortana. He told me that these examples reflected a philosophy that’s becoming core to how Microsoft sees the service. “We don’t subscribe to a model that the assistant needs to be all-knowing,” he said. Cortana is designed to be super-helpful “not by absorbing things, but by using experts.” It’s a distinctly different approach from Siri (which involves a lot of functionality hand-crafted for the service) and Google Now (which, above all, leverages Google’s vast knowledge of the world and the data it has about its users).