Now learn to speak to computers

It was in the mid-eighties when I was using a heavy duty Remington typewriter pounding out stories of how computers were going to replace typewriters. As we wrote about the coming computer age, the analogy used was the horse-and-buggy getting replaced with automobiles and true enough our beloved typewriters were soon junked to make way for PCs.

It seems just the other day, when the first computers came to our desks and things altered almost overnight. What didn’t change was the age-old Q W E R T Y keyboard. It was still the same way we typed on the keyboards of PCs, laptops, tablets and even smartphones as we did in the age of the typewriters.

However, even that’s about to be transformed soon, as the end of typing draws near with the advent of voice-controlled devices.

According to surveys, in 2017, there will be 24.5 million voice-controlled devices shipped, leading to a total footprint of 33 million voice-first devices in circulation. This is a large and growing market, and both Google Home and Amazon Echo, the first of the voice devices, poised to be huge successes.

The use of voice-activated digital assistant devices will grow almost 130% this year, according to eMarketer. Google predicts that use of these technologies will cause more than half of the queries to be voice searched by 2020.

For news publishers specifically, Voicelabs, a research organisation, found that 25% of Amazon Alexa users use voice to learn about news.

The year 2016 was sort of a watershed in voice controlled devices with Amazon’s artificial intelligence powered digital assistants Alexa and Echo gaining a huge acceptance among the application developer and common user communities both.

Strategically, Amazon heavily invested in tools for developing applications for Alexa. This led to a large volume of new applications, with news, trivia and educational categories dominating in terms of the number of applications.

Not to be left behind, Microsoft announced that it has partnered with audio equipment maker Harman Kardon to put the tech giant's voice-activated assistant, Cortana, into a smart speaker of its own.

Image and voice recognition technologies are becoming so sophisticated that experts predict both technologies could replace typing, at least in part, when it comes to communication and search. This in turn is going to throw up fresh challenges or opportunities, depending on how you view it. Few people type the way they speak or the way they use images to communicate, which means that web publishers will have to manage their search-engine-optimisation (SEO) strategies to accommodate picture and voice-related search queries.

As these technologies advance, consumers will also have to learn how to interact with them more fluidly or "learn how to speak to robots." From a business perspective, voice and pictures can be easier to analyse for personalisation than text, creating smarter sales and customer service experiences.

Image-based platforms, like Snapchat filters and Google Lens, use machine learning or real-time, user-generated images that can replace typing locations, temperatures, store hours or other details.

Advances in machine learning and artificial intelligence are also making it possible for apps and websites to infer information about images and show relevant details without users having to type to search for more info.

Nearly 70% of voice requests are natural/conversational language. New technologies, like Twilio's natural language understanding software are learning how to analyse intent during any voice call, which will make it easier for humans and voice-powered robots to have more authentic conversations.

On the image side, a new app called Project Chalk has been launched that will let users draw on top of live video chats through augmented reality, which could partially replace chat bot interfaces for interactions like customer service.

As we give up typing, we need to learn how to speak; how to speak in a way that machines can understand us. That’s a new area of technology that is opening up when anyone can give a command to a computer in natural language and the machine will translate that into a code without needing a programmer to write it.

But computers will also have to be trained in Natural Language Processing (NLP), a field of Artificial Intelligence, to create the ability to understand human speech.

As I write this I tried to train my Apple Mac to recognise my voice to key in a few lines, and I tell you it’s tough speaking to the machine.

The days of typing are soon going to be a thing of the past as voice commands take over. But I will continue to miss that delightful clickety-clack of my Remington typewriter keys.