I alluded to this idea last December, and it seems Larry Page has been thinking along the same lines – unsurprisingly, as Google is his company. I’m sticking with my prediction of 2013, but he seems to be hinting he thinks it will be sooner than that.
To paraphrase my previous post:
So far, you’ve seen computers in schools as things for kids to do spreadsheets or, gawdhelpus, Powerpoint presentations on. A combined typewriter and adding machine. But in fact, you already use one supercomputer regularly; it’s called Google. In future you will use many more without even thinking about it. And just how super do I mean? According to futurologist Ray Kurzweil, we can expect computer intelligence to match our own by 2013 (see graph) – just six years from now.
Try this thought-experiment. Thinks of fairly complex web search, the sort where you are not confident Google would give you the answers you need. Now imagine that Google gets a program that would allow it to write back to you and ask questions about your search, to understand it better. Plausible? Good. Now imagine that over four or five years, Google refine this program to the point where you can’t really tell whether it’s the program asking you questions, or another human being. Still plausible? Well guess what, your imaginary Google has just passed the Turing Test, the best-established benchmark for human-like intelligence in computers.
My takeaway from Larry’s presentation:
…there is less information in your DNA (about 600 MB) than in a modern operating system.
I suspect that is debatable as we still know very little about how DNA compresses information, but it is still a great start point for thinking about the issue.