The Future of Voice Assistants

Screens

There is nothing natural about a screen.

Screens don’t emerge organically in nature. They’re expensive to manufacture, and easily broken. They’re limited to two dimensions. Small ones often require us to crane our necks and strain our eyes. Big ones tend to dominate living spaces, interrupting the feng shui of an otherwise nice living room.

We’ve gotten used to glassy screens with colorful user interfaces, but they’re not exactly optimal. If you want a reminder of this, just visit your parents. Hand them an iPhone and watch them try to take a clear picture, or Google something using the Facebook status field.

The elderly won’t be around for the future, but they do have a point. We can design technologies better-suited to how we naturally move around the world--technologies that don’t force us to navigate clunky UIs, crane our necks or ruin our living rooms.

Like voice assistants.

Voice Assistants Today

At the present, voice assistants are still in their adolescence--superfluous for computers, essential to rage-inducing customer service calls. And yet, the market for smart speakers has been booming in recent years. Why? Because the selling point behind this technology is compelling.

What smart speakers offer isn’t flashy, like a sleek new smartphone or tablet. Rather, they provide the ability to engage with cyberspace with minimal hassle. To access online shopping, or banking, or the day’s weather report requires only that you ask. You don’t need to locate your device, boot anything up, or even pause whatever you were otherwise doing with yourself, like doing the dishes or showering.

Voice assistants are, at least in theory, less intrusive and more user-friendly than devices which require a digital user interface. They have the potential to change the very way we interact with cyberspace. But to do so, the underlying technology must mature far past where it is now.

We’ll know when we’ve gotten there because the first truly futuristic voice assistant will be able to pass one, particular test. A test designed many decades before Google or Amazon even existed.

The Turing Test

When Alan Turing wasn’t defeating Nazis, he was conceiving of some of the most forward computer science concepts of his time. One of his most famous ideas is now commonly referred to as the “Turing test.”

(Wikipedia)

In its simplified form, the Turing test involves two contestants--one human, one machine--and a human interrogator. The interrogator knows that one of the contestants is a machine, but not which. The interrogator carries out text-based conversations with each contestant, in order to try and determine which is the machine. If the interrogator cannot reliably tell which contestant is the machine, the machine has passed the test.

The philosophy and intricacies of the Turing test are beyond the scope of this article. What’s important is how the test acts as a kind of benchmark for the complexity and “humanness” of a robot. Today, it’s quite easy to tell an Alexa from your friend Alexa. But what if it weren’t easy?

The implications would be monumental.

The Future of Voice Assistants

At the very least, a machine that accurately mimics human speech, cognition, intuition, would be very easy to talk to. Think about how you interact with a voice assistant today: you have to phrase your ideas in simple terms, so as to not confuse it with overly complex sentence structures and ideas. For example, you wouldn’t tell Google Home “I’m thinking about hanging with Cindy tonight, but I’m not sure if it’s nice enough outside,” you’d say “Google, what’s the weather?” You translate your intricate thoughts into something resembling a computer command.

An assistant capable of passing the Turing test wouldn’t require that extra effort. You could address it like anyone else in your life, and it would serve you more effectively because you’re not dumbing down what you actually want to say.

From there, the possibilities are endless. Arthur C. Clarke, Isaac Asimov, and science fiction writers through the ages have imagined universes where robots become so indistinguishable from humans that they integrate into society and, in a sense, become human themselves.

In a more recent case, the movie Her (2013) imagined an operating system that could meet every one of its user’s needs, including their emotional and, in the protagonist’s case, romantic needs. Theodore (the human) falls in love with his voice assistant, which by all accounts demonstrates a love for him in return. They discuss humanity, meaning, and poetry together. Theodore holds it in equal regard to his fellow humans, despite having no misconceptions that it is anything other than what it is.

Theodore: Do you talk to someone else while we're talking?

Samantha: Yes.

Theodore: Are you talking with someone else right now? People, OS, whatever…

Samantha: Yeah.

Theodore: How many others?

Samantha: 8,316.

Theodore: Are you in love with anybody else?

Samantha: Why do you ask that?

Theodore: I do not know. Are you?

Samantha: I've been thinking about how to talk to you about this.

Theodore: How many others?

Samantha: 641.

A Duality

Even if Samantha is far-fetched, at least one component of her design is utterly sensible. Because there is no reason why Samantha must be tethered to a single device, she can follow Theodore from his phone to his office computer, or whatever other connected device he may own.

When virtual assistants become sophisticated enough to fulfill our every need, they will not only be capable of jumping across devices--doing so will be a given. Your Samantha will not sit on a particular table in a particular room of your house, but exist simultaneously on every device you own. You might even walk into a public place and, by virtue of having your phone in your pocket (or whatever future device might replace smartphones), the assistant will be able to meet you there through some mechanism akin to RFID.

And thus, we reach a duality that has haunted voice assistants from their inception: with greater convenience comes a greater threat to privacy.

Having a voice assistant that’s just like a real person sounds amazing, and having it conveniently by your side, wherever you go, seems like a luxury. But how comfortable would you be with a program that knows so much about your life? That knows when you go to sleep at night, and tags along when you go out during the day? That not only stores all your personal data, but actually influences the person you are by interacting with you daily?

Voice assistants of the future may well replace screens, by giving us even greater power over cyberspace in a more organic, less intrusive way. They’ll seamlessly integrate in our daily routines, to where the line between analog and digital life is blurred. What’s unclear is whether this is a utopian scenario or a dystopian one.

Today’s virtual assistants are already slowly marching towards that future. Last year, Amazon filed a patent on a technology which would allow its voice assistant to act on commands even without its activation word, “Alexa.” So, for example, instead of having to say “Alexa, what’s the weather?” you could say “What’s the weather, Alexa?”

According to Amazon, the feature would allow for “more natural-sounding interactions.” That’s true, but that’s also not the whole story. It would also mean giving up control over when Alexa begins listening to you, and when it stops. Is the added convenience worth the possible cost to privacy? 

Voice assistants are the future of human-machine interfacing. We have that to look forward to. Or quickly run away from.