Teaching computers how to think

Picasso and the singularity

I stumbled upon this great Picasso quote the other day. It’s pretty deep.

“Computers are useless. They can only give you answers.” — Pablo Picasso

It struck me that this might be the easiest way to explain the singularity — that theoretical moment in computing when machines outpace human intelligence.

You see, computing is still very much a one way street. We feed questions into computers, and out come answers on the other end.

But we’ll likely reach a day where we’ve fed enough into computers that the tables flip. A day where computers become powerful enough to start asking questions on our behalf. Or better yet, start asking questions to themselves.

After all, if a computer can ask a question and that computers are networked, it means computers can question each other. With billions of devices and counting, that sounds like a pretty powerful thing.

Clayton Christensen once said that “questions are places in our minds where answers fit”. Really, what that means, is that questions are a necessary first stop in our perpetual quest for answers. It also means that if computers can somehow ask questions that we’re incapable of asking (and it seems to me likely that they could — since questions are fundamentally about connecting formerly disparate data points) we’ll also be a step closer to finding answers that we were previously incapable of finding.

And one day, when they’re asking all those questions, maybe, just maybe, we can do the computing on their behalf.

Then again, what do I know? No answers. Just questions.