What Is Interaction?

In Crawford’s The Art of Interactive Design, he posits that interactivity is a process between two actors, each able to process the other’s actions and react accordingly. He notes three factors in interactivity: listening, thinking, and speaking. Each actor must be able to perform these actions in turn.

I’m not sure I buy this. Crawford notes that there are degrees of interactivity, and what he deems to be “interaction” is that which, basically, seems interesting enough to an adult to hold their attention. The example he gives is of a fridge door opening and causing a light to turn on. This might be interactive to a child, giving them hours of endless fun, but to an adult it seems too simple to be considered interactive.

However, if the example of interactivity for adults is assumed to be a website, I guess I’m just curious about where we draw the line. A computer program is, at its core, a system of ones and zeroes — hit switch, get light repeated about a million times. Obviously this is a greater degree of complexity, but at what point do we draw the line to say the non-human actor in an interactive scenario is genuinely thinking? A computer program is still, on a much more massive level, following some sort of decision tree — it can’t actually mimic the process of thought.

Maybe with artificial intelligence we can reach the model of truly “listening, thinking, and speaking” in interaction, but I’m skeptical that this model works, and “I know it when I see it” doesn’t seem a satisfying answer to the question either. I might posit that there’s some number (not one that I currently have) of potential outcomes from inputs to a device that leads it to be interactive, but this seems unsatisfying too.