In under a decade, software developers predict that “voice assistants” like Apple’s Siri, Amazon’s Alexa and Microsoft’s Cortana will be capable of speaking with us in natural, flowing conversation – guiding our Internet search queries, the function of our homes, and the direction of our very thoughts and imaginations. Recent dramatic technological progress in areas like speech and image recognition has allowed for rapid improvement in their ability to complete tasks and connect us to the goods and services we need. Now more than ever before, these interactive, speech-driven AI helpers are able to make life easier in every direction. And the more our machines learn about us, the easier it will be for them to intuit what we’re thinking – perhaps even before we’ve reckoned with the consequences ourselves.
For over a century, visions of the future have contained images of enormous screens used for video calls and “smart homes” that could respond to us using voice. Home automation has long been the cradle of invention for machine learning interfaces, where today their service can be integrated into everything from refrigerators to cars.
When you speak to Alexa, or Siri, or Cortana, or Google Now, they provide hands-free access to everything online and within the Cloud. You no longer have to scroll or click; all your requests are fulfilled via voice command. According to this website, consumers use voice assistants in their home more than anywhere else – at 43%. Using hands-free voice assistants on the road while driving ranked second at 36%. A widening scope of devices – lighting systems, music programs, televisions, home security technology and more – are able to work with personal voice assistants, bringing a multitude of home services together to create the possibility of a total smart home environment.
Growth and Competition
Even when conversations were misconstrued, early voice-driven intelligence users were eager to speak to Siri, despite her limitations. Then and now, she doesn’t always hear directives correctly the first time around. Her flippant answers seldom deviate from her programmers’ script and belie her ignorance. Fortunately, if you tire of Siri, you can choose to chat with her competition.
Google Now is at your service, and Facebook’s M, though still in development, is also an emerging contender in the space. Microsoft recently introduced Cortana, and Amazon is behind the popular Alexa who, among other places, lives in the “Echo”, a speaker device compatible with numerous home products and software programs. These increasingly-sophisticated helpers proactively listen to your every word. But while they fulfill requests, give directions, and create other reminders, they also retain a “memory” record of every operation.
Learning from the Mirror
The Internet has quickly evolved to become both a ubiquitous experience, and one that is increasingly “personalized” – thanks in no small part to the ambitious A.I. algorithms driving advances in natural language and machine learning systems.
Current virtual assistants reflect our desires, but do little to stir or instigate new patterns of thought. The A.I. programs working today are able to predict “preferences” by analyzing repetitive actions and previous data activity. As such, there are serious concerns that relying upon the intelligence of a machine will inevitably work against us to reduce our capacity for spontaneity, limiting our knowledge in the longer term.
Talk to your assistant, and you add to the growing treasure trove of user information – expanding her understanding of you, as well as where you fit within larger subsets of population data. Virtual assistant “brains” live in the cloud. Rather than relying on programmed, call-and-response type conversations of the past, voice-driven softwares interact using flexible code to better understand larger groups of people at once. To better synthesize knowledge from spoken language, Alexa and other helper bots use massive artificial neural networks to “train” and learn how to make more accurate inferences. These enormous networks depend upon unthinkably large amounts of data, but with users inputting millions of pieces of information every second of the day, it’s not necessarily in short supply. Complex beyond the bounds of human capacity, machine analysis techniques are now crucial to the process itself. Alongside Amazon, Apple and Google are also immersed in deep-learning research.
“At each increase of knowledge, as well as on the contrivance of every new tool, human labor becomes abridged,” wrote Charles Babbage, creator of the first mechanical computer. Voice-assistants are moving us towards new breakthroughs in efficiency, helping us outsource mundane daily tasks and to-do’s, and coming ever closer to comprehending our more complex wants and needs. But as we reshape our world around this new technology, it’s hard to predict just what we’ll lose in translation.
Guest Author Bio
Kate Lindsay is a writer and blogger based in the Windy City. Fueled by coffee and chocolate, she’s an MSU alum with a passion for recycling and refurbishing old furniture. Her favorite Girl Scout Cookie is the trefoil.