The uncanny valley is coming. Perhaps it’s already here. And I’m not sure anyone is ready.
Now, before you go and dive into the fallout shelter or zombie bunker, let me explain where I’m coming from.
First, let”s define terms: the ‘uncanny valley’ is a term from the realm of robotics that is used to describe a series of phenomena experienced when dealing with almost life-like human analogs. It specifically refers to an unsettling feeling that accompanies interaction with an almost perfect humanoid robot. They key to this concept is ‘almost’ – that is, there are things about the robot that are extremely lifelike, but also things that are off or incorrect in some way. Sometimes these things are fairly obvious – a strange eye twitch, an unpredictable motion pattern, or a speech flaw. But many times, the feeling accompanies much less tangible features. A common example from pop culture is the animated film The Polar Express, which featured animated characters made using motion capture from real actors. This early use of motion capture with animation created characters that seemed off to many viewers, giving the film a somewhat creepy vibe.
I imagine you are all now saying, “but Dustin, I’m not in an animated movie, how does this effect me?” Well, as is common in new media realms, the term uncanny valley has now begun to describe a broader range of things, modern technology included. I would like to briefly address what I see as ‘The Uncanny Valley of HCI’ – a trend that will affect all of us relatively soon.
So, we now know what an uncanny valley is, now let’s apply that idea to HCI (that is, human computer interaction, or how people interact with technologies today). Now, we all are constantly interacting with technology every day – from cell phones and computers, to iPads and projectors. And we’ve developed some fairly robust methods for interacting with and controlling them. There’s the tried and true buttons and switches, the classic keyboard, with which I am writing this very post, and the recent newcomer, the touch screen. All of these things have become standard ways to interact with our technologies. They weren’t always this way – the first mouse connected to a computer was an extremely revolutionary idea, but we have come to seem these methods as some of the most efficient means for interaction today.
However, times are changing once again. Technology has advanced to the point where previously impossible means of interaction can now be implemented reliably. From voice and speech recognition to motion and gesture control, we are reaching a point where we can use most of our physiology to control our technology. And that’s awesome, right?
Well, I am going to argue that for a duration, we will enter a pretty extreme uncanny valley because of these new technologies. For instance, I can tell my phone basically anything, and it will do it’s best to figure out and complete my request. Pretty awesome, right? The tech of the future. Or maybe not. I have encountered two problems with this tech – one which can be solved, ironically, with better tech, and one broader, more open question.
The first is the inability of the system to complete my requested task, reliably, more efficiently than I could just do it myself. Text entry is so fast, and the UI for my phone so streamlined, that there is seldom a scenario where one key phrase communicates my actions as quickly, and most importantly accurately. Now, this is a problem that tech will come to solve – speech recognition will get better, processors will get faster, and I will learn how to best query audio based UIs in time. However, this does not solve a second, more overarching issue – I’m talking, aloud, to my phone.
Well, duh, of course I am. That’s what speech control is, right? Well, yes, but consider this: the last time you took a call in public, how did everyone around you react? Were they happy? Excited? Annoyed? Upset? I wasn’t there, but I would guess thrilled wasn’t on the list. People naturally feel annoyed by the business and intrusion of others, an issue compounded by the one sided nature of phone communications. Now, imagine, for a moment, every person on the street speaking to their phones, about any manner of things – sending messages to friends, scheduling dinner, taking selfies – all messages spoken aloud. Pretty annoying, I’d wager.
And yet, one of the most fascinating devices currently being developed as ‘the next leap in personal computing’, Google’s ‘Glass’ wearable computer uses audio as a core piece to it’s operation. Living in the area of the nation I do, I see people using Glass somewhat regularly, and it always feels like they are muttering things under their breath, likely trying to wrangle some function out of their new toy. Now, I know Glass is in beta, and certainly not in a final form, but I can’t help but wonder what happens when we all have Glass (or some other device) on our face – how does that effect social norms? Is it suddenly ok to just sort of mumble through life, absorbed in my own computing world?
And this is just one isolated technology. We are now beginning to see gesture based systems, where users merely wave their hands to control digital interfaces, virtual reality headsets and experiences, where users feel fully immersed in a simulated environment, and even sense based systems to share smells with your tweets. The realm of possibilities for what is possible for people to interact with their technology has exploded, and there is a lot of new tech coming down the pipe.
In the end, I mostly wanted to say that I am really excited to see where these new technologies will lead, but I am also wary that the market will be flooded with a lot of new and different tech, and not all of it will work as well as we’d like. Until norms, standards, and even social conventions get caught up, there will be a lot of weird and ineffective systems being tested in the market, which will likely make for a rocky few years in HCI. So buckle up, and get ready to try some new and exciting tech – just keep an open mind and you’ll be all set.
Welcome to the Uncanny Valley. Population: All.