Tony Poulos

Voice is not dead

As featured in DisruptiveViews 

So, voice is supposed to be dead (or at least dying), according to the leading telecoms companies in the world. Supporting this theory are countless reports on Gen X, Y and Millennials espousing that they don’t talk to each other, preferring to communicate by text messaging.

It seems both ‘assumptions’ may be a little premature and that voice is not only not dead but making a comeback. People do actually talk because it is a natural way of communicating and doesn’t need any tools to be effective. When you want to communicate something you simply open your mouth and out come the words. What could be easier?

Mary Meeker, general partner at Kleiner Perkins Caulfield & Byers and author of the highly respected ‘Internet Trends’ report, is also a great believer in the future of voice. She states that voice should become the most efficient form of computing input, largely because it is hands and vision-free. “Voice lends itself to an ‘always on’ way of life.”

“Humans can speak 150 words per minute, for instance, but can type only 40 words per minute. The conversational aspect of the medium lends itself to personalized experiences with computers understanding context from previous questions the user has asked and the user’s location.” While many voice recognition tools can be frustrating to use now, Meeker said that when speech recognition reaches 99% accuracy, people will go from barely using the tool to using it all the time.

“Speech recognition accuracy rose to about 90% in 2016 from about 70% in 2010. And the use of voice has been rising noticeably. Google voice search queries, for example, are up 35 times since 2008. Sales of voice-based devices such as Amazon Echo could be just about to take off, compared to more text-dominated devices such as the iPhone, whose sales peaked in 2015.”

Ms Meeker may be on to something here. Many new model cars have voice-activation capability built-in for doing anything from turning on the radio to setting destinations in their GPS systems. You will soon be able to drive your autonomous motor vehicle with voice commands.

Voice activated assistants are now standard fare on most smartphone operating systems and more and more apps are using voice to operate. Instantaneous translation apps are able to take one voice stream as input and output the same stream in any number of chosen languages. They even compensate for accents! And ‘chat bots’ with voice recognition and response are becoming prevalent on customer self-care sites.

Banks are introducing voice recognition systems as a secure means of access. Voice patterns, it seems, are as unique as finger prints but don’t need special readers to be used, just a microphone.

Yes, I get the theory and I like the convenience but I am still having problems talking to a machine, any machine and I suspect I am not the only one. Is there anything more stupid than someone walking down the street saying to their iPhone “Hi Siri, I am lost, can you tell me where I am?” or asking their smart watch what time it is?

Setting aside the embarrassment of talking to machines in public for a moment, talking out loud could prove disastrous in a hostage situation or if you happened to be close to where a crime is being committed. Will they let you use voice commands on a plane? And what happens if you have laryngitis?

That all pales in significance if you assume (and we must assume) that voice conversations will be picked up ‘accidentally’ by any one of your devices. For example, you might be watching the football and get annoyed by a decision that goes against your team and you shout out, as you do, “kill the ref”! Your digital assistant, not knowing it as a harmless figure of speech might unintentionally set off a series of actions that may not only be embarrassing but also costly. I, for one, can’t wait to hear those horror stories.