BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Get Ready For A Future That's More Siri-ous

This article is more than 7 years old.

“Hey, Siri.”

With those two little words, we’ve ushered in an era of voice recognition. Sure, anyone who has an iPhone knows that Siri - like her Microsoft counterpart, Cortana - has a distance to travel before the interaction is as seamless and useful as one might wish it to be. Still, that it works as well as it does is a triumph and it’s getting better all the the time. More importantly, we’ll look back on the Apple voice recognition system as the beginning of it all: the dawn of the voice-originated Internet.

In her deck this month, Mary Meeker noted that person-to-machine interactions are going to skyrocket. Right now, speech recognition accuracy is at 95 percent -which sounds like a lot but to her point, not many people use it. Seriously, how often do you ping Siri for an answer to an idle question? Meeker says that when the accuracy jumps to 99 percent - just four measly percentage points - adoption will radically increase, especially as the ability to comprehend context and semantic nuance improves. Likewise, she notes, voice recognition must also solve the latency problem. Technology has, in some sense, spoiled us - and even 10 seconds of waiting feels like something of an eternity. Too long and we’re on to something that can give us what we need, precisely when we want it.

Interactions are likely to focus in the areas you might well expect: local information, fun and entertainment, personal assistant, and general information. More broadly, voice will supplant how we currently ask machines these questions: experts put the number of queries that will take place with voice recognition at 50 percent in only 4 years. This is not remotely outlandish or fanciful thinking. Consider the benefits to pinpoint-precise voice recognition. Driving and not sure where to go? Open your mouth and ask. Thinking about a new coding technique? Use VR to access Stack Overflow. Walking around a new city and want to find the best bistro? Ask your phone. It’s like Joaquin Phoenix in Her - without the creepy factor. It’s going to make being out and about, whether at home or on travel, much, much easier.

More practically for technologists, this means the pressure is amped up for building great product with great VR capability, where it makes sense. Think everything from cars that can answer you with a projected diagnostic when you say, “Hey, Prius, what’s that noise?” to design software that changes the color gradient of an image by a specific percentage to travel apps that can integrate location data with reviews to offer up restaurant recommendations on the fly. But remember: incorporating voice recognition only makes sense if you can really deliver at that 99 percent accuracy level - and that’s only if VR is really needed for your product to begin with. Anything less than that and - to Meeker’s point - it’s not prime for widespread consumer adoption.

As voice recognition becomes more accurate, and thus more relevant to the everyday person, we’re going to shift from a highly haptic and manual Internet to a voice-originated one. It’s going to utterly change our habits. Interactions will largely be at the speed of our own thoughts; the primary limiting factor will be how fast we can get those questions and desires out of our mouths. The way we will then experience the Internet will be a compelling blend of visuals with voice. As entrepreneurs and investors, the opening up of the Internet in this way will mean a series of fascinating opportunities as we ask ourselves:

  • What’s the role of technology when interactions are no longer haptic?
  • How will voice liberate us as users?
  • How can we use voice to surprise and delight users - whether they’re consumers, an enterprise customer, a small business, or nonprofit

This is the chance to be at the forefront of big, radical change that's certainly coming.