I just checked this book out of the library and will probably be a bit busy for the next few days engrossed in it.

When I was in college, I was excited by the idea of artificial intelligence (AI). I imagined incredibly patient and creative tutors helping people learn whatever they wanted to; robotic scientists conjecturing new theories based on all of the available evidence and proposing and conducting the best (most efficient) experiments to grow our knowledge in important areas as quickly as possible (thus solving the most pressing problems such as diseases, energy, food production); wonderful companions for the lonely (and everyone else who might be interested); automated software engineers to build customized high-quality solutions for people in an interative process (since most people are unable to fully specify what they want until after they get what they ask for a few times…); artificial artists creating pictures, films, books, and music to entertain and enrich us; etc…

I don’t believe there’s anything magical about people that enables us to perform intellectual feats that are impossible for computers to do, in theory. We are, after all, machines.

I read and loved Gödel, Escher, Bach, of course, and was encouraged by the thought that lots of very smart people were working on these problems.

Unfortunately, progress has been slower than I hoped. There have been many advances, and we have some “clever” programs that do cool things, but not the kind of things I had been anticipating. Over the years I’ve been casually following progress in the area, as well as prospects in biotechnology and nanotechnology. All have great potential, but as time has gone on, I’ve been lowering my expectations about when we’ll see real world-changing progress.

Ray Kurzweil knows a lot about this stuff, and has a pretty good record at predicting technological progress. It’s not possible to precisely predict things that depend on individual discoveries, but it seems possible to recognize the nature of the rates of growth and to find bounds to when things can reasonably be expected. According to Kurzweil, we’re approaching the “knee of the curve” of exponential growth in these areas and can expect to see phenomenal progress in the next few decades.

I appreciate Kurzweil’s optimistic view of these changes. Many, like Bill Joy, tend to focus on the dangers and urge us to stop progress. But I don’t think it’s wise, or possible, to put the genie back into the bottle. Knowledge can’t (and really shouldn’t) be destroyed. If we leave it to others to develop it, we’ll only be at greater risk. Bad guys will always have the advantage that destroying is easier than creating. Our best chance lies not in hiding from progress, but in learning as much as we can about the dangers so that we can devise counter-measures as quickly and as well as possible.

In any case, while I’m somewhat familiar with some of the information in Kurzweil’s new book, I’m excited about seeing how he puts it all together, learning a lot more about some details, and knowing when he thinks certain developments are likely to occur.

Even if he’s wrong by a few decades, it’s going to be a wild ride and we’ll probably start seeing a lot of cool stuff relatively soon.

It’s a great time to be alive.