The concept of personality has us hooked; just look at Cosmo quizzes and the thousands of online personality tests. And rightly so: it’s something that has profound effects on our friendships, love lives (that old “she’s got a nice personality” chestnut) and careers. For instance, Bruce Tognazzini claims that designers must have an ‘N’ in their MBTI, one of the slightly less dubious profiling tools. (I actually agree with him on this. I’m an INTJ myself.)

However, we’re also a little infatuated with personality, and often assume that someone’s actions are caused by the ‘type’ of person they are, while ignoring the social and environmental forces that influence them (the fundamental attribution error). In reality, personality is always framed and affected by the world around us, meaning behaviour can be quite variable. Just because someone’s angry once, it doesn’t necessarily mean they’re an angry person. We have to work backwards, interpolating someone’s underlying personality from several observations of their behaviour. You can’t really get to know someone from a minute in their company.

For instance, at a football match, I drink, swear, and slip into a latent Welsh accent. This is no surprise—my environment almost demands it of me, since I’m surrounded by drunken, sweary Welshmen. But you’ll find me behave very differently in bed with a girl, going through airport security, or talking to my Nan. This behavioural variance is part of being human and people who lack it are deemed to be boring. If you behave the same in a nightclub as in a library, you won’t be invited out again.

Constrast this with technology, which behaves in a very rigid manner—the same in all environments. I think it’s time to make technology more interesting by introducing some mild behavioural variance. Sampled over a few readings, we can then start to form an opinion about the underlying personality, which is where we make those emotional connections.

Clearly we can’t go too far. Some behavioural consistency is essential for usability, and some devices are better suited to quirkiness than others. However, the dead zero we’re at now is clinical and drab.

Fortunately, we have the jigsaw pieces we need to imbue technology with personality. We just need to put them together. As mentioned above, behavioural variance generally comes from environmental influence. This meshes nicely with technology’s increasing context-awareness. Bluetooth, RFID, APIs, accelerometers, spimes etc, common geek parlance, all refer to ways technology is becoming more aware of itself, other technologies and us. But it doesn’t need to be this esoteric. Glade recently released a quite silly air freshener that only activates in the presence of a human.

The concept of an emotional response to technology isn’t new, by any means. For example, the uncanny valley:

I happen to think the uncanny valley is bullshit, but I challenge anyone to watch the following and not be slightly saddened:

So let’s imagine an operating system that sees you’ve split up with your girlfriend and says sorry. A program that knows you were out drinking last night and therefore uses muted colours and suggests you take frequent breaks. A mobile that loves going on rollercoasters.

This could be so much more fun. And the exciting part is I don’t think it’s too far out of our reach—for starters, we already give out plenty of these informational cues (knowingly or not):


Ultimately what we’re aiming for is intelligence (or at least pretence thereof) in technology. In the words of Piaget, “intelligence is the ability of an organism to adapt to a change”. I think behavioural variance is a perfect example of this adaptation, and for that reason I think we shouldn’t be scared of giving our future technology a personality of its own.

Based on my lightning talk “A rainy day, lost luggage and tangled Christmas tree lights” given at Skillswap On Speed, 29 Oct.