Paul Gilster: Better voice technology ushers in era of enhanced digital assistant

Voice is a supple user interface, the logical way to interact with shrinking devices that carry ever faster processing speed in our pockets. A computer that talks to you used to be something of a joke as we went through the early stages of voice technology, but increasingly adept artificial intelligence is going to change all that. Phones and other wearable gear won’t just talk; they’ll anticipate your needs by monitoring your life and putting the data you need in front of you.

Smartphone voice options such as Google Now and Siri are slowly getting better, and at moments can be almost charming. Of course they’re just programmed to respond and use scripts for various situations, but even so, it was fun to set a wake-up alarm on a road trip recently and have Siri respond: “Alarm set for 6 a.m. Don’t wake me up.” Canned speech or not, a talking smartphone with an attitude can seem like a cross between a digital assistant and a pet.

Edge to Google

Voice is a feature that is going to improve dramatically in the near term. Microsoft’s Cortana is the one thing that could drive Windows Phone sales out of their doldrums, and the company is already running effective ads showing dialogues between the Cortana voice and Siri. Even Blackberry, whose market woes are well known, is rebounding with a voice entry. Its new handset, called Passport, will come armed with a personal assistant app that talks to you.

I’d say Google Now has the edge when it comes to providing messages in context – Google has gotten quite good at this – but as we increasingly integrate voice with our phones, we’ll see a scramble between the big players trying to use voice for a greater range of input, with Apple said to be exploring ways to use IBM’s Watson in ways that Siri couldn’t dream of today. A new Google-commissioned study shows that 55 percent of under-18 smartphone users activate voice search at least once a day, while 41 percent of adults do the same.

Better controller

My impression is that Google sees clear sky ahead in voice technology and wants to build user interest by creating studies such as this one, which was handled by Northstar Research. Oddly, 56 percent of adults in the study say that using voice makes them feel “tech savvy,” which surprises me because conventional search requires more work than running a voice search. I tend to use Siri mostly for making phone calls, something the study cites as common practice among teens, while Google reports that the average adult uses voice search to get directions.

The next generation of digital assistants will use voice for more than launching apps or making calls. After all, we’ve moved into the era of cloud computing and enhanced machine learning. As smartphones get wider data pipes to the cloud, they can interact with the heavy-duty processors needed for complicated voice interactions. A mobile device wants to be minimalistic – you don’t want to have to perform a series of actions when a single voice command can do the job. Voice also makes a far better controller for household gadgets such as TV remotes. Who wants to show? Smartphones and wearable computers have no more natural interface than the human voice.

As to listening, expect phones that anticipate our requests, popping up maps without anything more than a mention in a conversation (the phone as eavesdropper), opening up message fields for quick voice data entry, finding Web pages relevant to a location. Phones that talk and, even more, phones that listen, are a bit creepy until we figure out where the best uses are and where the device is intrusive. But the era of the enhanced digital assistant has already begun.

Paul A. Gilster is the author of several books on technology. Reach him at