The Cognitive Umbrella

Originally published on Tumblr.

Cognition is complicated. Perceiving, understanding, and knowing are hard to define. Philosophers since Aristotle have been trying to develop frameworks to help us understand understanding. And now, without a firm grip on human cognition, we’re developing synthetic equivalents.

Current artificial efforts are primitive, but they’re already starting to have a significant impact. It won’t be long before cognitive computing makes it onto 60 Minutes, if it hasn’t already. After beating all of the human Jeopardy champions, IBM’s Watson — the poster child of artificial cognition — was featured on the cover of magazines and newspapers throughout the world.

So why now? Why, after 2,500 years, is cognition going mainstream. Three reasons:

Marketing.
Fear.
Money.

The first one’s easy. Aristotle didn’t have a marketing department. Watson does.

Fear is understandable. Until very recently, perception, understanding, and knowing were the features that made humans special. Now that machines are “catching up”, we’re on the verge of being knocked off that pedestal. That this will affect our thinking just as profoundly as the realization that the universe didn’t revolve around the Earth. Fear isn’t a surprising response.

Cognitive computing is going to be extraordinarily valuable. Until now, software has made knowledge workers more productive. We’re now starting to realize that it’s going to be able to replace a great many of them. For the software’s owners, there’s a lot of money to be made. For everyone else, fear isn’t a surprising response.

When cognitive computing started becoming popular in software circles, I was annoyed. I was annoyed by the imprecision. I was annoyed by the hype. I was annoyed because cognitive computing wasn’t anything more than an umbrella term covering a great many “traditional” areas of science and technology. I’m no longer annoyed.

Cognition is fuzzy by nature. There’s a reason philosophers have been trying to get there arms around it for so long. The hype is justified. This family of technologies will make our world unrecognizable. And finally, having a framework in which to place technologies from artificial vision, through adaptive interfaces, to language processing, is useful. Each of these on their own solves a small set of problems. Together they will allow artificial beings to perceive, reason, and act like us. Only faster.