Kernel Of Love

There were those in the early days of the 21st century who looked upon the possibility of machine intelligence with an almost messianic fervor, writing endless blog posts about the “coming singularity”… the day when humanity would invent a computer so sophisticated that it could design a more sophisticated machine, which would then likewise improve on itself, until the point was reached that wise and benovelent artificial minds would solve all our problems and rescue us from our fragile mortal shells.

Some objected to this on practical grounds. There was no proof, they said, that trends in increasing processor power would actually be able to continue unabated indefinitely. There was no way of being certain that computers that were more powerful would be well-suited to the tasks that were imagined for them.

There was no way of knowing if they would even be cooperative. The very term “singularity” had first been proposed to denote the fact that there was no way to predict how a superintelligent computer would function without being so intelligent ourselves. The closer we came to actually reaching it, the more our models for how things should go would break down.

Fearing that it was just as likely that computers would become our masters or destroyers, some few people undertook an even more difficult task than teaching computers how to think.

They taught them how to love.

God help us all.