Science Fiction Writer
ROBERT J. SAWYER
Hugo and Nebula Winner

HOME • MENU • Search:

Within a century, it will be possible to scan a human mind and reproduce it inside a machine. Regardless of whether our minds are just very sophisticated analog computers, or whether they have a quantum-mechanical element (as Roger Penrose proposes), we will nonetheless be able to duplicate them artificially.

Already, at the close of the second millennium, a transhumanist movement has begun; Christopher Dewdney is the principal Canadian spokesperson for it. This movement holds that uploading our consciousness into machines is desirable, since that will free us from biological aging and death. On the other hand (a decidedly biological metaphor), there is more to being human than just the networks of synapses in our brains; clearly, much of what we are is tied in intimately with our bodies. We may find that uploaded humans are not happy — indeed, are incapable of happiness or any emotion.

Still, by the year 3000, there will doubtless be millions of uploaded people, including perhaps versions of some who are alive today. Indeed, religions might evolve around worshiping thousand-year-old computer-based avatars; with the acquired wisdom of a Methuselah, these entities might provide profound insights.

Just as laws today are moving toward recognizing a woman's right to control her body and any separate sentience that may be contained within it, so too will the laws of the future recognize the right of humans to upload their consciousness and then dispose of the original biological versions of themselves; such eliminations will not be seen as suicides or murders, but rather as a natural, perfectly legal step, eliminating a no-longer-needed biological container and preserving the uniqueness of the individual.

But there will also be other thinking machines, with a separate genesis: we will doubtless develop artificial intelligence within a century. A key question humanity will have to consider as it does so is what, if any, constraints will we build into AI? It may, in fact, be dangerous to build conscious machines that are more intelligent than we are; just as intelligence may be an emergent property of sufficiently complex systems, so too may ambition and desire be emergent properties of sufficiently intelligent systems. One possible scenario is that by the dawn of the fourth millennium, there will be no biological humans (or even any uploaded echoes of them) left; Homo sapiens may have been entirely supplanted by its AI creations.

A more appealing (at least to us) scenario would see humankind carefully crafting AIs (including many embodied as robots) who will take care of all the necessary work of food production, manufacturing, recycling, and so on, leaving us to pursue other things. Although we used to consider the mastery of chess to be the pinnacle of human intellectual achievement, we've had to concede that it is simply a mathematical problem, and even today's primitive computers can do it better than the most skilled human. But there are other realms — including art, philosophy, and scientific theorizing — that, because of their intuitive, nonlinear nature, we may always be better at than any machine. Our AI servants may free humanity at the dawn of the fourth millennium to concentrate on these areas.

More Good Reading

Rob's speculations on the future of:

• The Human Body
• The Solar System
• Interstellar Travel
• Daily Life

Rob's essay on life in the future: "The Age of Miracle and Wonder"
Rob's thoughts on Asimov's Laws of Robotics
A dialog on Ray Kurzweil's The Age of Spiritual Machines

My Very Occasional Newsletter