When Google unveiled its artificial intelligence (AI)-powered Google Assistant in October 2016, CEO Sundar Pichai commented the world is moving from being “mobile first” to “AI first”.

The prediction already appears to be coming true. In addition to AI-powered digital assistants including Apple’s Siri, Google Assistant, Microsoft’s Cortana and Samsung’s Bixby featuring in smartphones, some AIs are also making inroads into consumers’ homes in the form of connected speakers.

Amazon led the way with its Amazon Echo range running its Alexa AI (though, ironically, the latest device in the range features a 7-inch touch screen). Google quickly followed with Google Home, Harman Kardon is preparing to launch the Cortana-powered Invoke, and Apple unveiled its HomePod speaker at its annual developer conference on 5 June, which will be launched in the US, UK and Australia in December.

The emergence of voice AIs in devices designed to take pride of place in users’ homes poses an obvious question: are smartphones losing their position as consumers’ central interface?

Jonathan Collins, research director at ABI Research covering smart homes (pictured, left), told Mobile World Live (MWL) the smartphone’s position is certainly being challenged, but not because of any “waning in the value of the smartphone interface.”

The analyst explained smartphones “will remain central in most locations,” but will “sit alongside other interfaces” in consumers’ homes.

Smartphones’ role in the home is evolving from being the central to an additional interface because: “aspects of the device design that enable it to be mobile and personalised – rechargeable, easily portable, a mandatory companion – do not suit a role as sole smart home controller,” Collins said.

David McQueen, an ABI Research analyst with long experience covering the mobile device market (pictured, right), agreed smartphones will complement home-centric connected speaker products: “I tend to believe that voice AI gets used more in the home (or in a car) than on smartphones owing to it being positioned in a more private, exclusive environment.”

While McQueen believes smartphones’ position as the sole interface in the home may be changing, he predicted the mobile devices “will still be used for other voice AI applications, notably navigation”.

Francisco Almeida, senior research analyst covering European Mobile Devices at IDC EMEA, agrees the device used will vary by task.

While Almeida (pictured, left) conceded “theoretically, smartphones would see their position challenged at some point,” he noted consumers have been slow to use existing AI-based assistants in their daily lives: “There are tasks they could already be doing with Google Assistant or Siri, and they don’t see it brings the required value compared to touch, compare to their status quo.”

In the long term, voice AIs could be the latest evolution in user interfaces, following the shift from pushing buttons to touch screens, Almeida explained. However “there is still a lot of education to be done on the consumer side” regarding the benefits and limitations of AI technology, he said.

Platform land grab
Another aspect to consider is whether the smartphone as a device is losing its place as the sole focal point for platform providers.

McQueen highlighted most of the connected speaker products available today use voice AIs already available for smartphones, noting Samsung is yet to push Bixby beyond smartphones and predicting other vendors are likely to develop “their own voice AI platforms”.

For Collins, companies “have seized the potential to leverage their smartphone platforms” as part of a broader “land grab taking place for control in the smart home”. At this stage, speakers are allowing the delivery of existing platforms “in a more suitable form factor.”

Such an approach, though, surely presents something of a problem for platform providers, which have built a large ecosystem around their current smartphone-oriented systems?

Almeida agrees: “There’s still no ecosystem synergy, so even early adopters that are technology savvy, they might have an iPhone as their smartphone, an Amazon Alexa at home, and use Cortana in their work laptop. So there is not the required interconnection of data to take this to the next level of utility from a user’s standpoint,” he told MWL.

Talking proper
As recent reports of Samsung’s struggles to train Bixby in the intricacies of English, and Apple’s decision to launch HomePod in three English-speaking markets attest, language remains a key obstacle for voice AI platforms.

With so many players developing their own systems, McQueen queried whether the market for home devices could become fragmented “with various voice AI platforms requiring specific cues (wake words) to set them off?” There are also questions over content curation: “For example, if you ask Alexa for a particular Beatles song to play, do you care where it gets the song from and who handles that decision as long as it gets played?”

Almeida noted Amazon is being methodical in its approach to deploying Echo to ensure the devices fully understand all possible variations – accents – in individual languages.

In time, Almeida speculated, the ability to cover a wide range of languages and accents could become a point of differentiation for voice AI platforms and, indeed, overcome a key weakness of the current market: namely that consumers still choose a device based on its features and price rather than the AI it runs.

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.