It’s So Nice Talking to You, Mrs. Robot.

Living in an Age of Digital Assistance.

In 1983, a song line was on everyone’s lips “Domo Arigato, Mr. Roboto.” It was the refrain of Mr. Roboto from Styx and one of the verses starts with, “I am not a robot without emotions, I’m not what you see; I’ve come to help you with your problems, so you can be free”.

While this song is rarely played these days, the utopia it was describing has, in some ways, become true. We have become very comfortable talking to machines. And the main reason is that they understand what we say and hence get us to the right solution. Gone are the days of frustrating menus in a voice response system or the hesitation to just “tell me how I can help you” because we doubted that the machine actually could.

We now live in the age of Siri, Alexa, Hey Google, Erica, and so many more speech assistants. Most of them are female, so it is not really Mr. Roboto solving our problems, but Mrs. Roboto.

Research shows that a growing number of consumers now actually prefer talking to a chatbot over a human because it gives them more control and there is a felt relief from “being sold something I don’t really want or need”. Of course, that notion is not really true because as we are interacting with machines, highly targeted advertising is all around us on the web, utilizing everything we have said or done online, but that is not the focus of today’s column.

I want to take a look ahead at how Mr. or Mrs. Roboto will further evolve and become so interwoven into our existence that we will soon not be able to imagine our lives without them. We will look at improving basic capabilities, then personalization, and finally humanization as three major drivers.

Basic Capabilities Are Improving Dramatically

The technology disciplines we are looking at are called Natural Language Understanding and Natural Language Generation. Machines can understand us because technology can precisely recognize what we say and recent improvements are allowing the machine to correctly conclude what we mean by what we say and hence what we want. The error rates are going down and our comfort level goes up. Isn’t that the same thing we look for in other people? We just want someone who understands us… And the machine can respond in a very human-like manner, not in a robotic tone.

Most commercial chatbots (whether they are text-based or voice-based) still have a limited range of services they can master. The main reason is not the ability to understand, but a limitation in the backend information they can access. They were designed to do a few things for the customer.

But as Siri and Alexa are showing us, if you define the backend more broadly (i.e. the entire internet), then there are few limits to what they can find. My five-year-old loves talking to Siri, asking her all sorts of questions, and I am sometimes stunned at the precision of her answers and the resources she points him to (and no, he does not have his own device, it is still supervised learning).

Increasingly Personalized Robots to Match Our Preference

There are a few things you can do to current chatbots, like pick a different voice, but all in all we accept that these machines are “anonymous”, and that they represent a general technological capability rather than something that is deeply personal to us. But that will soon change. Here is what it will take.

First of all, we currently cannot “trust” a chatbot. We know that Apple or Google have access to all of our information and recent revelations of employees “listening in” to their customers are deeply disturbing. Someone recently told me “if something is free, then you are the product”. Our actions, our behaviors, our search requests are data, and that data has value if it can be collected, analyzed and sold. But some smart company will soon figure out that people are ready to pay for protection of privacy while still enjoying the technological capabilities.

Once we can trust that what we share with our virtual assistant will not be shared with the world, we will be ready to make “our robot” more personal. We will assign personality characteristics, such as different styles of humor (by design, Siri can be a bit sarcastic at times, while Alexa is more no nonsense). We may also give it a visual representation, from simple avatars to more complex human representations (see next section).

As our comfort level grows, the machine will “get to know us”. We will download all aspects of our personal history that we have digitally available (professional, family, social media, maybe even financial or health, depending on the level of trust). It will learn our preferences and habits, and it will be able to make even more precise suggestions and recommendations tailored entirely to our needs — than today’s bots already do.

It will also be able to look out for us in terms of eating right, exercising, or not spending too much money. Apps that do that already exist, but they are single-purpose apps. We will not need those any longer once our trusted personal advisor is with us every step of the way.

Some of us will prefer a physical presence. It may be a cuddly companion toy for our kids, a pet-like robot that can talk to us (way beyond Tamagotchi), or a humanoid robot that can move around the house. We will enjoy the ability to switch between physical and virtual representations because we cannot easily take the actual machine with us, but we still want to be able to have access.

Humanizing Robots — The Final Frontier

At this point, we are close to having a “companion”. Someone who understands us is always available to talk to, looks out for us, anticipates what we need or want next, and points us in the right direction (literally and figuratively). We may have assigned personality attributes that are highly compatible with our preferred styles. Our companion can be funny or serious, motivates us with challenges or never argues, keeps us engaged, or stays quiet when asked to.

It can be a gateway out of isolation and loneliness into connectedness with others, or it can be our best friend in solitude if we so choose. It is so much easier to avoid the complexities of human interactions if you can change the settings.

The movie Her (2013 by Spike Jonze) may have been among the first to tell the love story between man and machine and it was solely based on falling in love with a “personality”. Now imagine we could give our companion a face, a body, or a style of fashion, again with unlimited variations available to create a perfect match. We would probably stop calling it “it” as I have done throughout this article, and start referring to “him” or “her”, if we have not already crossed that line much earlier.

Samsung introduced Neon Life earlier this year, the ability to create hyperrealistic computer-generated humans or “artificial humans”. We will be able to design our perfect virtual companion. And the choices in physical representations may before long include very human-like forms. Synthetic skin over robotic skeletons.

Hidden Dangers?

How can real-life possibly keep up with that sense of perfection? Will we still value differences and idiosyncrasies in other people? Will we accept little faults and be willing to look past bigger ones? Or will we become so selective that finding real friends will be harder and harder? Well, maybe we don’t have to worry about that, because we already have the perfect friend.

What will happen to love and intimacy? What will happen to birth rates if more people choose to live in “blended” relationships? I know, this may seem a bit far stretched, but technological progress is so fast that it may leave us little time to think about these things in advance. The fact is that our comfort level is already shifting. Domo Arigato Mr. Roboto may catch up with us sooner than we think.