When you say “Hey Alexa, what’s the news?” your smart speaker can now repeat Yle’s latest news broadcast. All you need is the Flash Briefing news app and Yle as your news source.
This pioneering feature was built in five days as part of Yle Beta‘s Alexa project, with Lassi Seppälä of Yle and Tuomas Lahti of Qvik as the programmers.
“No one knows for sure yet what voice-controlled user interfaces will be like, so it is fascinating to build this kind of service at such an early stage”, Lahti says. “The technical side of development is quick and easy, and it’s great to be able to build a working prototype in just a couple of days. There is still some work to do on UI design, though.”
In the United States, smart speakers made a real breakthrough last year. More than 37 million voice-controlled smart speakers have been sold, and one in six US households is estimated to own one. Voice-controlled services are not used, let alone developed, in any significant amounts in Finland as of yet, and Yle’s news-reading feature is one of the first Finnish Alexa services.
The world’s tech giants are in a race to invest in the development of voice-controlled devices. In addition to the Google Home speaker, Google offers the Google Assistant application that can also be connected to the devices of other manufacturers, such as smart TVs. The battle between Alexa and Google Assistant was one of the hottest topics at the annual CES exhibition of innovative technology. For now, Alexa is much more widely used than Google Home. Even though the Apple HomePod is entering the stage a bit late, it will probably be a tough competitor for the two.
Voice UIs require new ways of thinking
Your alexas and googlehomes differ from more common voice-controlled services like Siri in that they are genuinely built for voice control, not just physical UIs that can also be talked to. Siri is used by an estimated 375 million people worldwide each month, and 20 percent of mobile Google searches were voice-controlled in a survey conducted some years ago. As smart speakers are mostly used in the privacy of your own home, the threshold for talking to your devices can get even lower.
“Onboarding and affordances are vital because these are still deterministic systems that rely on input triggers”, says Jesse Heys, Product Designer at Qvik. There is still work to do in this regard, since no less than 72 percent of smart-speaker owners admit to not being familiar with all the functions of their speakers or able to use all of them.
“Text and graphics are more forgiving, even when conventions are ignored, because the possibility space of input triggers is comparatively low. Likewise, there aren’t many different ways to interact with the triggers.” Voice and language is less forgiving because the possibility space of input triggers is dramatically higher, more opaque, and more complicated.
“The technology required to interpret intent and meaning from all the different ways people can try to interact with smart speakers just isn’t advanced enough yet”, Heys says. “Right now, voice UI feels like learning spells.”
In the spirit of the Yle Beta program, the project for bringing Yle news to Alexa was a quick overview of the technical aspects of developing voice-controlled features. The UI was kept simple with the aim of getting a serviceable first version out to see what kinds of voice UIs are worth offering to users.
“When the designer is not tied to a physical screen, you can approach service design from a whole new angle”, says Qvik’s Head of New Business Development, Pia Höglund. “The service development scene is just starting to take shape, and we have a great opportunity to be involved in figuring out what you can do with this stuff and the best way to do it.”