Hmm! I guess we do that all the time. In this case, I’m talking about the future of the PC. Or data processing in general. Whatever.
I read this story about Google automatically identifying and blurring faces, and it reminded me of the year 1980.
In 1980 I was a graduate student at Andrews University in Michigan, and there was a computer show. It was pretty small, as befitted the state of the industry at the time, but of course I was there to see what was already there and what was coming. For some reason a friend of mine and I got into a discussion with one of the vendors about the future of computer science. In particular, we both felt that in time, a computer would be able to drive a car, given the appropriate sensors.
“Impossible,” he said. In particular, he stated that object recognition would be just too difficult to accomplish.
And there’s why this seemingly unrelated article made me remember the story. A face in a crowd is much harder to distinguish than the objects we were talking about–signs, trees, and so forth. Yet the recognition is here. In fact, since we were talking about specially equipping the road with distinctive markings and even transponders, what we were talking about would be considerably easier.
Of course, in reality robotic cars already exist, even if only in prototypes. That would be more to the point. But still, object recognition is a key element.
I now return you to the relevant parts of life, the universe, and everything.