Networks are supreme agents of change, in my humble opinion. Case in point: the institution of mail service in the 16th century provided the network that Locke and Galileo used to exchange ideas. Mail service was enough of a network to spark change that would shortly see almost every European institution change radically after 15 centuries of stasis.
Thus, it would seem that big, fast networks can cause gigantic change at a tremendous pace. And, although our global network now reaches 100 million people, this is still less than two per cent of the world's population - which suggests that things have only just started to get moving.
So, on the eve of the millennium's last year, I'm prepared to offer a glimpse of the our world in 20 years (still recognisable), and 50 years hence (maybe not!):
The world in 2020
Band width will be cheap and ubiquitous. The fibre optic infrastructure will expand and upgradable optical switches will appear, removing the roadblock of constantly needing to replace expensive switching hardware to increase capacity.
Wireless connectivity will exist nearly everywhere. Almost everyone's "last mile" connectivity will be via wireless loops. Phones and appliances won't have jacks on them - they'll have embedded antennae or transducers instead. "Leapfrog" societies, like China, will mainly have wireless infrastructure, obviating the slower process of stringing wire from poles. Growth industry in "old" Western nations will be telephone pole removal.
Computers will disappear: even computer scientists won't use computers, at least not the big boxy things we're used to in 1999. Computing and networking resources will be ubiquitous - built into walls, floors and vehicles and clothes.
Very small machines - key ring-sized mobile phone/ personal digital assistant (PDA) hybrids - will enable us to command these resources for image projection, computing and communication. Really small machines, christened "smart dust", will fill connectivity gaps between devices and media.
We'll be able to conduct meetings by asking our PDAs to get our colleagues on the line. Smart software agents will ask other agents to track them down and pull them together in a secure teleconference that will display on the nearest wall.
For $1,000 you will be able to buy computing power equivalent to a human brain, but true human-like machine intelligence still lies in the future. Machines will begin to do a lot of "productivity" work, like drafting routine business e-mail and conducting transactions, and these programs will train themselves by watching our behaviour, or the behaviour of other machines.
Routine customer service calls and e-mail will be more often handled by software than humans, and it will be hard - but not impossible - to tell the difference. Semi-autonomous robots will appear and do useful work like ditch digging (for more fibre optic cables) and maintenance. Growth industry: robotic supervisors - humans who will, for insurance liability reasons, stand around while robots do the work.
Cyberspace will begin to merge with "real" space: spectacles and cameras will use embedded Global Positioning System circuitry and wireless database access to overlay data on to live scenes. For example, your sunglasses will mark out buildings of historic importance, and display directions for walking or driving - and so will cameras, binoculars and tiny video cameras.
The world in 2050
Governments will start to disappear. Elections on the Net will lead to the discovery that governments are increasingly useless and inefficient.
We'll need publicly supported infrastructure, but private companies will deliver it far more efficiently than increasingly self-absorbed and isolated governments. The Internet's potential exploded when it went commercial, and so will functions once reserved for government. Imagine government services that are good, cheap and easy to access.
Bureaucratic institutions will be to the new world what monarchies are to our own - largely useless, if interesting, anachronisms. This is not to say that people will suffer. People will be freer and richer than they are today. The only roles left to governing bodies will be setting standards and arbitration in the case of disputes.
In a highly interdependent world, military force will be all but useless except in the case of primitive rogue states and possibly some equally primitive terrorist groups. Information warfare will be the primary focus of the world's militaries, and they will be highly dependent on private companies.
The trend to decentralisation of power will accelerate. Boundaries will come to mean "distinctions between interest groups" and not "imaginary lines on the earth".
Fully autonomous robots will do a lot of the work that makes the world wealthy and smooth-running. Robots will routinely have the intelligence of cats and dogs, and do jobs such as delivery, construction and maintenance. Maids and butlers will make a comeback in the form of tireless robots.
The first sentient machine will appear: with perhaps a Turing rating of .45. The machine's software will have been developed with techniques borrowed from biology. The machine will learn and be trained rather than programmed in the current sense.
The first autonomous, self-replicating space probes will be on the drawing board: privately financed, 1,000-year missions will begin to probe the galaxy. The SETI program will have long since changed its name to SET- T (the Search for ExtraTerrestrial Technology), and will be on the threshold of being able to detect advanced communications technologies.
Don't believe me? Then follow the links below.
Berkeley Endeavour Expedition
JPL Adaptable Hardware for Evolvable Computing
MIT Laboratory for Computer Science
MIT Raw Project
The Age of Spiritual Machines by Ray Kurzweil
The Space Frontiers Foundation
Robert Anton Wilson
Hans Moravec: Transcendent robots
Search for Extra Terrestrial Intelligence