If you imagine a graph with the degree of realism a robot attains running across the bottom and with the amount of positive empathy it can generate climbing vertically, then, generally, you see a positive correlation between the two, and an upward-sloping graph.
As the robot gets more human, we like it more. However, the Uncanny Valley hypothesis suggests that there's a point where this breaks down; that there comes a moment when the increasing realism in something that's still obviously a robot doesn't create feelings of empathy and connection, it freaks us out.
That's one side of the valley. The hypothesis further suggests that as robots get yet more realistic, then this effect diminishes - we effectively forget that it's a robot and empathy returns. That's the other side. Except no-one knows this, because no-one's ever built a robot that good.
Why am I telling you this? Because advertising will enter its own Uncanny Valley in the next few years, and I'd like us to be prepared.
The online advertising industry took a step closer to its Holy Grail last week with the introduction of various advertising initiatives on Facebook, MySpace and across other social networks. These promised much better targeting than has ever been possible before, massively increased relevance. Advertising that's so in tune with its audience that it ceases to be advertising and becomes simple, useful information.
Hmm. Sort of. There's certainly better targeting, or at least different ways of targeting, but, as yet, no-one's going to feel like they're reading something specially designed for them. It seems like we're about to enter a period where our digital lives will be full of the online equivalents of those messages you find on your television when you check into a hotel: always welcoming someone who's got a name a bit like yours. Never actually your name. And you wish they just hadn't bothered, you wish they'd just issued a general, warm welcome and not tried to connect at a level they just didn't really feel (because if they had really felt it, they would have made sure they had gotten your name right).
This online marketing revolution is going to generate quite a lot of these creepy feelings. We're going to be wondering how companies know so much about us, why they're talking to us in such a familiar way and how come they get everything just slightly wrong. At this point, we might find ourselves responding more favourably to those brands and advertisers that can master the compelling generalisation and the universal truth. We might remember that great communicators can connect with millions by knowing only one thing about us, that we're all people.