|Its an exciting time to be a computer engineer or scientist. Things are really about to get interesting. Many authors (eg Ray Kurzweil, Hans Moravec, Verner Vinge, Eleizer Yudkowski, and Damien Broderick) predict that computers and robots will reach human level performance by around 2050. The interesting thing is it won't stop there. Computers will quickly surpass humans and be many times improved just a short time beyond 2050 (or whenever they actually hit this milestone). There will also likely come a time when computers will be able to evolve independently at a rate many billions of times faster than ordinary biological organisms. Once this starts to happen computers will begin to program themselves, and have minds of their own, separate from human programming. "The Terminator" is one pessimistic view of what might happen next.
Well before mid-century, I predict that virtual worlds within computers will become so compelling, that people will spend there liesure time more with computers than in the real world. This is already the case for many of us, but it will eventually be the case for nearly everybody - like it or not. I'm not saying this is a good thing, just what I expect to happen. Ultimately we may eventually download our consciousness into the machine and live out whole ordinary lifetimes in the span of a micro-second, and in the process become virtually immortal and maybe a little bored. Perhaps we are all living in a computer "matrix" of some past advanced civilization right now and not even reallizing it.
The biggest problem with this view is that no one has any idea how to make a computer have consciousness. We know very little about what biological consciousness is, and much less about how we can make a computer experience it. So even if we make computers with a zillion times more computational power than a human brain, we may not be able to make it conscious. The problem is not like computer chess (or go) where we know that we are all the time making incremental progress toward beating human players.
If we are going to reach this advanced state of technology within the next 40 years, what will life be like 400 years from now? That is very hard to imagine. In fact there is really no way for us to predict what happens after technology "spikes" sometime during the next century. This is a major theme of "The Spike" by Damien Broderick. If there is still any life at that point, it will most likely be very different from what we experience today. The point is we are currently at the knee of a super-exponential curve, and the rate of change is about to increase dramatically. Ray Kurzweil noted that during the next 100 years we will experience 10,000 years of progress at todays rate. Conversely, the last 100 years have only given us 25 years of technological progress at tadays rate of change.
My biggest worry is that the rapid change will bring about an increase in the frequency of catastrophes as everything becomes more volitile. Some of the ways in which we could wipe ourselves out are: nuclear armagedon, new deadly strains of diseases and viruses, large environmental disasters, terrorist plots, computer viruses, and nanotechnology run amuck. However, in spite of the potential risks, computers and technology provide us the tools to manager these risks better.
The figure below (from Morevec's Robot) shows the super-exponential nature of computer power over time. Note that the graph has an exponential rather than linear shape even though plotted using a log scale for the vertical axis. By the time my son is 23 he will be able to by a computer that is 1,000,000 times faster that those available today for a fraction of the cost.
|Thoughts on the Future|