People make tools, and tools then remake social systems--that has been the case for millenniums. The heavy plow fostered the manorial system of governance in northern Europe in the 9th century, allowing for the accumulation of surplus, population growth and urbanization. The automobile gave us freedom to travel and commute, and in the process hollowed out cities and created suburbs.

How will computers remake today's social systems? No one can say for sure. Certain major elements of change--from how schoolchildren do research for papers to how and where we spend our working hours--are already apparent, of course. But the important issue facing us is this: Do we accept the consequences passively, or try to anticipate and control them? I would go so far as to suggest that our leaders should begin to restructure social, economic and government systems to accommodate the forthcoming change.

For all of human history, society has been built predominantly on a physical foundation of tangible objects and assets. Governments drew their strength from geography: Mountains, rivers and oceans protected them and defined their reach. Businesses depended on their physical assets--factories, inventory, natural resources. Communities were clustered around churches, schools and places of public business. Factories provided places to work. Streets supported the movement of goods.

Physical infrastructure serves a second important purpose: It acts as an information storage and transmission network. Office buildings facilitate communication between co-workers. File cabinets store information. Streets carry information in the form of people, letters, newspapers.

But for the past several centuries, the importance of society's physical foundations has been gradually eroding as intangible assets such as knowledge, organizational structures, methods of production, intellectual property and trade secrets have become more and more essential. As people developed more accurate and sophisticated methods of navigation and gained knowledge of world markets, countries were no longer dependent on raw materials located entirely within their borders. Improved methods of communication by telegraph, telephone and over data networks have practically eliminated the need for manufacturers to be located near cities.

As a consequence, cities are no longer needed as centers of production. Today, information technology is turning the gradual displacement of physical assets by intangible ones into a rout. NASDAQ, the New York Stock Exchange's most important competitor, does not even have a trading floor. It exists only electronically--and thus, in a sense, exists nowhere at all.

Reducing the cost of information transactions, and severing the bonds between information itself and the physical means of transporting and storing it, has affected almost every aspect of civilization. Gradually, the influence of government will be further circumscribed and the influence of corporations will expand proportionately. Unchecked by traditional forces, large virtual communities--those that unite their members electronically--will develop significant power over the direction of public policy. Tax systems that depended on physical constraints will wither.

Let's look at the transformations that are already underway:

* Special interests without accountability. Local "special interest" groups--bound by common religion, ethnicity or political goals--are a fundamental part of democratic society. They add diversity, foster bonds, preserve cultural and religious values and protect the interests of their members. When they act to further their own interests at the expense of others, they are subject to local forces that cause them to moderate their stances: resistant neighbors, editorials in the newspaper, opposition at the polls.

But virtual communities--free of many of these local constraints and founded not so much on the basis of religion or ethnicity as on other narrowly focused interests--are expanding rapidly, and information technology will amplify their power dramatically. A national or international special interest group is very different from a local group supporting the exact same causes. An organized virtual community that collects only a few dollars in dues from each of its millions of members is a very potent force, and is substantially less sensitive to the interests and feelings of others.

It took years to build up the National Rifle Association and the American Association of Retired Persons by means of pre-Information Age procedures. These organizations were subject to the constraints of the physical world--sending actual mail, holding actual conferences, maintaining local chapters in real towns and cities. Freed from such constraints, special interest groups can be expected to achieve critical mass sooner and become larger and more reactive.

One way to confront this problem would be to return more political authority to smaller political jurisdictions. For example, the federal government could let the states control all agricultural subsidies. This would make it more difficult for a virtual community of tobacco farmers to get northern states, for example, to subsidize tobacco production. The federal government should also consider getting out of the education business entirely, and letting states and local school districts set the education agenda. This would in all likelihood curb the ability of virtual political and religious organizations to influence the educational agenda on a nationwide basis.

* Corporations beyond control. The evolving information technology of the past century has given multinational companies the power to control and coordinate activities around the world. It has enabled the matching of the production of goods around the globe to worldwide demand, and facilitated the design of products in one location and their production in another.

Even when "information technology" consisted mainly of the telephone, multinational corporations became more difficult for governments to control. The companies could structure their operations to minimize taxes by moving profits to geographic locales that provided favorable treatment. When regulations increased costs, they could easily move production to other locations. But the advent of computers has enhanced these capabilities incalculably, and brought them within reach of even modest companies.

Not too long ago, governments thought they could control their national economies. They believed they could increase exports by devaluing their currencies and create jobs through deficit spending. Today, much of that power has been taken over by multinationals. If one considered multinational companies to be economies, 50 of them would be listed in the top 100 of the world's largest economies.

These companies have the ability to decide where the jobs will be and who exports and who imports. They determine the trade balance in the industry segments they control. By choosing the countries in which to locate, they decide, in effect, what laws they will obey. They have the flexibility to employ child labor, to pay very low wages, and to select locations that are willing to trade environmental damage for jobs.

National governments that wish to manage the situation should consider attempting to control the actions of multinational corporate "governments" through agreements rather than just laws, as the ordinary regime of national laws and regulations will become less and less effective.

* The collapse of taxation. Information technology poses a grave threat to the tax system, which depends not only on an individual's honesty, but also on the constraints imposed on those individuals and corporations by physical infrastructure. One of the fundamental requirements for a government's tax schemes is a knowledge of where money is earned, where a transaction takes place and where value is added. A second requirement is the ability to exert authority over those activities.

Information technology undermines each of these preconditions. Suppose a customer in the United States orders a product from a company located in a tax haven in the Caribbean and uses funds in an offshore bank to pay for the purchase. The company in the Caribbean sends an electronic message to its factory in Mexico to manufacture the product and ship it to the customer. With a little bit of financial engineering, this transaction becomes nearly impossible to tax. It is not clear where the transaction takes place; in fact, it would be extremely difficult for a government to determine if a product had been ordered at all. The type of transaction just described has been possible for some decades, by phone and wire. But as corporations have become more far-flung and communications more sophisticated, the opportunity for such transactions has expanded enormously. While these problems are not entirely new, information technology increases their intensity.

Individuals, too, have new freedom to avoid taxes--both legally and illegally. In the past, many professionals tended to locate near the consumers of their services or near the sources of information critical to their practice. But today, many engineers, financial managers, lawyers and others have the flexibility to live and work in states that have little or no personal income tax while at the same time continuing to serve clients elsewhere.

In general, owing to information technology, income taxes collected on a national level, whether they be on individuals or corporations, will be the most difficult taxes to collect--they'll be lost in the nether world of the intangible society. As it becomes easier to avoid taxation and the risks of doing so decline, more organizations and individuals will likely choose this option.

The easiest taxes to collect will be the ones based on activities of consumers that simply must take place locally: Gasoline must be delivered and sold locally, water and electricity have to go to actual businesses and homes, gas and heating oil have to be delivered to the point of consumption. It would be difficult to avoid taxes on real estate, food, transportation, entertainment, hotel rooms and health care. Indeed, many of the taxes that will be easiest to collect are regressive.

Taxes on most forms of consumption fall most heavily on the poor who spend a disproportionate amount of their income on consumables and energy.

If a government wished to get ahead of this problem, it would begin openly and deliberately shifting the tax base to consumers while designing rebate systems to make such taxes less regressive. Dealing with this problem after corporations and professionals have fled the tax base by both legal and illegal means will make confronting it all the more difficult.

We can allow our civilization to accommodate itself, willy-nilly, to the evolutionary dictates of information technology, as determined by the expediency of the market, or we can attempt to find policies that will influence it. The winners in a civilization remade around computers will not be those who attempt to contain information technology. The winners will be those who invent the new structures of government, business and society in which technology is embedded.

William Davidow, a former senior vice president of Intel, is a general partner in Mohr, Davidow Ventures, a Silicon Valley venture capital firm.