hyalineskies

“Mesmerising,” one friend of mine said. “It’s so cute,” another IMed me. “Totally cute,” said another. Another friend had an awesome smile on her face when she saw a video of it. Even I, the usual skeptic of everything tech, smiled when I saw the little, bispherical object kicking it on screen to Spoon’s I Turn My Camera On. That little yellow object is Keepon, a rhythmic robot developed by Hideki Kozima and Marek Michalowski, created to study social interaction of humans and robots. While Keepon was built for the therapeutic purposes of autistic and other under-neurodeveloped children, Keepon has had the ability to mesmerise and capture the emotions of ten various friends that I’ve shown the robot to. Kozima and Michalowski wanted to build a robot that could display and evoke emotional states, and it is safe to say that they certainly have.

However, Keepon is a perfect display of something more than just a cute novelty: it is a standing example of the human propensity to attach mammalian personality traits to entirely inanimate objects. The action is deeply embedded in language and history: large ships are given human names and referred to as “she”, we consider our computers and automobiles to be in bad moods when they break, and we constantly seem to attribute emotions of anger or other negativity to devices when they don’t function as planned. We place emotion in the personalities of devices with their personalities dictated by engineering error or physical malfunction, while we constantly design the outward appearance of our electronic devices to be increasingly clean, mechanical and otherwise futuristic. Within the sterility of our Macbooks and iPods, we place emotional tidbits and interfaces: we plaster our desktops with pictures of loved ones, we hold data dear to our human lives, and in some cases we build personalities for ourselves online in social networks or massively-multiplayer games. The personal in personal computer is not simply in the sense of it being built for one user as opposed to an enterprise; our computers and electronic lives are permanently bonded to our lifestyles and ways of being. They are a part of our real-life persona.

There is a point, however, at which it appears that we are not simply working with our devices as mechanical tools in our lives; we repeatedly attempt to attach greater emotive states to something that’s exceptionally emotionless. We name our computers human names or give them names we’d usually give our pets; we get frustrated and feel attacked when Word crashes, taking all of our work with it. While we superficially accept the mechanical nature of human creation, we secretly attach emotive states to it in a way to possibly somehow relate to the labyrinthine structures of silicon-based semiconductors. It appears that, in the end, we’re building devices, feeling unhappy with the interaction we have with them, and then actively searching to replace things with emotive states to make them more human-friendly, more organic. We want the mechanical æsthetic and the organic interaction.

What, however, is perpetuating this type of interaction? What emotive states are we actually seeking? No one would want a computer that was always ill-tempered or one that shared some other generally awkward human emotion (would you enjoy it if your MacBook Pro felt artificially sexually aroused?) There appears to be a point at which we stop the anthropomorphisation of the device underscored in pop culture: In Star Wars, thousands more are attracted to the naïve, traditionally cute demeanor of R2D2 than the rather obnoxious, worrysome one of C3PO. We find the cold mechanics of the liquid-metal T-1000 despicable while we exonerate the heroism of the older, more human-flawed Schwarzenegger in Terminator 2. We hate the inhuman, machine-like coldness of the agents in The Matrix but enjoy the cozy sociability of The Oracle. While all of these juxtapositions are between fictional machines, we love the machines that are more like us: flawed, slow to evolve, and technologically able to express emotions of sympathy, happiness, and, in some cases, love.

Keeping it optimistic: positive interaction

Because of this, we seem to desire only positive emotive states from our machines, leaving out any type of despicable feelings such as anger or envy, although life would certainly be more interesting if Outlook really did hate you when it crashed. This quick display of happy emotion, although entirely contrived, makes us feel better when we interact with the machine. Such an emotive state is easily seen in Susan Kare’s anthropomorphic iconography for the classic Macintosh boot screen, aptly named Happy Mac and Sad Mac. In both cases, we get a quick, extremely human analogy to the way the computer is operating upon boot: in the case of the happy Mac, we know all is well with our machine; in the case of Sad Mac, we know that something needs to be done to “comfort” (i.e. repair) the machine to get the data we need.

Macintosh hardware takes this human feeling to extremes, as well: closing an iBook or other newer notebook and suspending its contents is referred to as “Sleep” mode both semantically and actively: the LED on the front of the computer slowly fades in and fades out, mimicking the patterns of slow human inhalation and exhalation during sleep cycles, although the LED’s smooth effect is entirely useless during the Mac’s sleep cycle. Meanwhile, however, as humans we notice this sleep effect much more intuitively than we would a little yellow light. It evokes an emotional response in us as we almost feel as if we should be keeping quiet as to not wake the computer.

In these two examples, we are seeking positive emotive states as an indicator to interaction and well-being in status through emotional expression from the machine, much as we would see from another human. We can tell if another human is happy or sad from the same gestures; we can tell a sleeping human from one that is actively awake by their breathing patterns. Building the emotional interactive state both links the computer’s state to an organic cognate as well as bonds us more emotionally with a rather cold object. In this case, we seek positive interaction from the machine; by having human characteristics, the machine, half-anthropomorphised, appears to really be more than a complex calculator. Considering the immense amount of trust we put in the machine’s ability to maintain our real-life information, the positive feelings we receive from having the machine appear sympathetic to the human condition comforts us and makes us place greater trust in an object that is really something neither to be trusted nor distrusted.

Staying optimistic: positive feedback

While the idea that we want “feel-good” computing is evident in status events like Happy Mac, we want positive feedback from our machines just as parents give children: we want some type of emotive state of approval from the device. Keepon, the robot described above, has this ability to show approval, and the emotive state of approval has appeared to have a great effect on children in an experimental setting. Of course, a robotic system isn’t required for this type of approval for us to recognise it as approval: instead, the pure illusion of attentiveness allows for us to feel this way about our machines. The response times of a machine when we perform tasks makes us feel as if those that are faster are not only helping our productivity, but also somehow make us feel emotionally reaffirmed that the machine is working with us, not against us.

The converse of this statement is what I had described above: when the machine fails, we emotionally tie this to negative feedback; the machine’s failure, in some way, is an insult to us, a program crash accusatory as if we could have somehow prevented it through our own interaction. Such an “emotive state” causes us to become further alienated from the device: we cross an invisible border in which we begin to recognise the machine as mechanical rather than emotional as we call out the machine’s subordinacy. In this case, we are offended much as if someone we thought was close to us somehow forgot our name: the impersonality of the machine gets underscored to a point at which the machine itself is responsible for rebuilding rapport with the user. The attentiveness and reliability of the machine is not simply something we wish for for the sake of our own productivity; they are the same qualities we look for in fellow humans.

Emotion isn’t technical

In terms of usability and interaction design, we are constantly bombarded with all of the things we should never do, an ever-expanding list of design abominations such as the use of <blink> and broken code that only works on one browser or another. People such as Jakob Nielsen make huge amounts of money simply telling us what it is that we’re doing wrong as interaction designers, however, the list we are given by usability experts is always a discrete set of things that went awry when we designed one page or another.

Ironically, however, as complex as we make interaction design, as complex as we make the design of our interfaces and scalable, degradeable code structures, the user really doesn’t care. What the user wants in the end is nothing more than a little love from the machine. In essence, we want to evoke positive emotions when using our interfaces, not negative ones. We simply want machines that seem to care. All of the discrete rules we’ve learned on top of this, from building accessible web pages to applications of Fitts’ Law, can almost always be distilled down to the emotive state.

Emotional users are passionate users

Meanwhile, those of us building software, making names for ourselves online, blogging, or otherwise publishing content to the Internet and greater technological sphere seem constantly interested in gaining that one next reader: we want the traffic for profit in the case of businesses; we want the traffic for personal gratification in the case of non-”pro” blogs such as this one. We are always looking for the best way to make our users passionate about our cause or follow our philosophies in hopes of getting them to return sometime, and we stop at nothing to try to improve both our own techniques as well as our properties and products to get to the point necessary to gain the next new consumers.

However, in all of our complexity, the issue is very simple in the end: a passionate user is simply an emotional one. The best user experiences are not modelled on mathematics or patterns; they seem to be, in some huge way, based upon that fuzzy logic of Stephen Colbert: the feeling of truthiness goes a long way in building rapport with users or readers. While in many cases we are able to model these gut instincts with formulas and interactive patterns, they’re not always perfect predictors. Sites like Hampsterdance [sic] have been largely successful in the past not because they had awesome design: instead, some emotion was evoked within its visitors. MySpace, while horribly ugly, offers a social portal to an array of tons of people and evokes all sorts of emotion in its users, drawing them closer to the site as they become more addicted to the emotional rollercoasters of human-human, not human-computer, interaction. The best design, then, really has nothing to do with our theory: instead, theory is just our guide to building something that is generally regarded as beautiful. The best design is simply emotional.