Technalysis Research
Previous USAToday Columns

March 7, 2017
It may be time for a tech tax

February 17, 2017
Apple's transformation from bear to bull

January 13, 2017
Voice-controlled devices shift tech industry

2016 USAToday Columns

2015 USAToday Columns

2014 USAToday Columns

USAToday Column

April 7, 2017
It's noisy in here! The coming problem of too many voice assistants

By Bob O'Donnell

FOSTER CITY, Calif. — Think about the real-world equivalent—multiple assistants at your beck and call. While it’s not a situation too many of us will ever find ourselves in, it doesn’t sound too bad, right?

Well, unfortunately, in the digital world, it’s about to become our own worst nightmare. Having multiple voice-based assistants trying to answer or respond to our queries could lead to a cacophony of unwanted information.

The problem is going to crop up in several different ways. First, several Android-based phones are going to include multiple voice-based assistants on their phones. Huawei’s Mate 9, for example, includes Amazon’s Alexa, and with future upgrades to Android, will have Google’s Assistant as well.

Similarly, the Galaxy S8 that Samsung unveiled will eventually have its newly-created Bixby assistant as well as the Google Assistant.

And that’s only on one device. Most people have multiple devices with assistants on them—Windows 10-based PCs with Cortana; iPhones, iPads, and Mac with Siri; Amazon Echos with Alexa; Google Homes with Google Assistant; and—well, you get the idea.

The irony is that the world of voice-based computing and personal assistants is clearly an exciting one, and one that’s just starting to show some of its promise. After early failed experiments with the original version of Siri, many people are starting to really enjoy the accuracy and usefulness of the latest generation assistants, particularly Amazon’s popular Alexa assistant.

Right now, these services are separated by the fact that you have to call out the appropriate name to get each assistant working. So, whatever name you call out before you request information or before you request a specific action be performed is the assistant that will respond to the request. But it’s not hard to imagine a time in the not-too-distant future when each of these services either lets you create your own “response” name—let’s be honest, that’s what most people want—or just responds automatically because it recognizes your voice. When that happens—and it will—you could end up with multiple devices listening and attempting to respond to your request all at once.

Initially, these multiple responses might be kind of intriguing and a good way to test which assistant is more accurate and provides the most useful reply. But after, oh, two or three times, it’s going to be a disaster, with each assistant barking out its response to your requests simultaneously, essentially destroying the entire process.

Worse, even if you learn to be specific about your requests, it will probably turn out that assistant 1 is better for one kind of information, assistant 2 is better for managing your calendar, assistant 3 is the best option for tapping into your files, and so on. Realistically, of, course, no one is going to be able to remember which is which, leading to reliance on a single assistant of choice.

The core assistant vendors, of course, would probably like to see that scenario, with each believing that they have the best mousetrap. Ironically, though, the best answers from each assistant should all be the same, which makes the challenge of trying to create differentiation that much harder.

As a result, we may end up needing to create a hierarchy of assistants, with one serving as the master and being intelligent enough to know when to pass off certain requests to other assistants. Unfortunately, the likelihood of that ever happening is pretty low, because it would require each of the assistant vendors to be willing to share certain critical details with their competitors and risk becoming a “subordinate” assistant.

Of course, even if one company does end up offering a demonstrably better assistant than others, it won’t really be useful unless it’s equally available across all the devices we own and use. That’s part of the reason we’re starting to see Amazon, for example, bring Alexa to iOS and Android. But even there, part of the reason Alexa works well on the Echo is because of the array microphone arrangement on the Echo—something you won’t necessarily find on other hardware products capable of running Alexa.

Despite my concerns, I’m a big believer in voice-based computing. Until we can start to get past some of these platform-centric limitations, however, I think we’re heading for some very challenging times with digital assistants.

Here’s a link to the original column:

USA TODAY columnist Bob O'Donnell is the president and chief analyst of TECHnalysis Research, a market research and consulting firm that provides strategic consulting and market research services to the technology industry and professional financial community. His clients are major technology firms including Microsoft, HP, Dell, and Qualcomm. You can follow him on Twitter @bobodtech.