Commentary

As one door closes, another opens. Today's announcement by the FVAP Experts Panel that the Pentagon's remote Internet voting project SERVE is inherently a security sieve and should be scrapped, probably marks the end of a ten-year flirtation with Internet voting that has finally come to nothing.

But running elections isn't the only thing that governments do. In fact, it's one of the things it does at most a few times a year, while the other functions of government take up much more of its time, and also its money.

Instead of using computers and the Internet to help out in elections, why not use them to help government do all its work, all the time. At a time when states, such as California, are strapped for cash, why not ask millions of ordinary computer users to donate their unused microprocessor cycles to a CaliGrid™, or other state, multi-state, local, or federal grid? But what exactly is "grid computing"?

To learn the basics and some advanced aspects of "grid computing," click here to hear an interview with Paul Kirchoff, Vice President of Marketing and Business Development at United Devices, a leading grid computing company.

Ever since 1996, I've been trying to figure out, then popularize, ways of using advanced technology to make government more responsive, more cost effective, and more democratic. These efforts have focused, successively, on Internet voting, Smart Initiatives, and e-government.

Of the three, e-government has obviously been the most generally-accepted and is now the most prevalent. Internet voting has had, and continues to have, a checkered existence, due to logical constraints, technical shortcomings, government intransigence, and political bickering. Smart Initiatives, the use of digital certificates on smart cards and other computing devices to digitally and remotely sign official petitions, such as those calling for initiatives, referendums, and recalls, have not yet even been seriously considered in the political space, but that may be changing soon.

I've argued, on and off, in support of wireless broadband networks as a way of bridging the digital divide and providing valuable and essential Internet connectivity for everyone, thereby eliminating the digital divide and facilitating the adoption of Smart Initiatives, Internet voting, and enhanced e-government services on a more equitable basis. Wireless technology seems a good way of getting the Internet to income-challenged individuals and families, domestically and especially in countries with no real electronic infrastructure to speak of in place yet.

But someone has to pay for the devices and those in power in many poor countries don't want their populations to have easy access to the business, cultural, political, and religious ideas of the entire planet. Besides, private wireless companies seem to be pushing a rapid adoption of their technology by everyone who can afford it without any help from the government as part of a program of social and political empowerment.

But recently I started thinking more about "grid" computing, a phenomenon that involves the linking, via the Internet, of large numbers of individual computers into "grids" that allow each device on the network to access the computing power created by the aggregation of all the microprocessors in all the linked machines, synergized through the use of grid-enabling software. By taking advantage of the "down time" or unused processing cycles on those machines not being used to their fullest potential, the collective grid network reaps the benefit of having a vast collection of circuits that can be put to use in complex and computation-intensive calculations.

Thus, meteorological institutes, DNA crackers, and star-gazing astronomers can use what is functionally a supercomputer for their work at a price vastly lower than what the equivalent supercomputer would cost. The Chinese, the Indians, and researchers in the U.S. are hurrying to build and use these grid systems to do their computing work faster, better, and more cheaply. The growth of these grids means that soon the most pressing issue involving them will be finding new and innovative ways to put their computing power to good use.

So why don't we hook up all the computing devices in city governments, state governments, and the federal government into grids to provide cities, states, and the national government with the power to do more, better, faster, and less expensive processing of their work? Why don't we virtualize more and more government functions, so that the collective processing power of the grids will have enough work to do? (We should also make arrangements to deal fairly with the government employees who depend on their work for their livelihood and who would be replaced under this scenario.)

Cities can link their grids, as can states. The eventual result will be networks capable of delivering government (cyber)services to millions in a rationalized, cost-effective, and standardized way, with individual cities and states free to customize the generalized power of the GovernmentGrid™ (or GovGrid™) to suit their own preferences and the desires of their constituents.

Government jurisdictions could export their excess processing capacity over the Web, in the same way that electric power companies do with electricity. India and China already "export" their excess programming capacity, just as Hollywood exports its films. Creating a global market in processing cycles would yield all the benefits in this area that free market advocates already argue accrue in regards to any product or service, including increased investment and the stabilization of supply and demand.

Private companies would be free to purchase processing power from GovGrid the way they already purchase water and electricity from pubic utilities. If a government decides it wants to subsidize the consumption of computer processing power among the poor, it could do so. And just as alternative energy producers are now free to feed their excess power into the electric grid, individuals and groups could be paid to feed their excess processing capacity into the GovGrid or other, private grids.

The result of all this giddy gridding will be a drastically-increased level of computing capability for all participants, the full use of the existing hardware/physical plant, and the enhancement of those functions that depend on the swift and accurate operations of computers.

It might even be possible to focus the power of GovGrid on the task of encrypting the data involved in e-government, and e-commerce to such an intense degree that it will become virtual impossible for any other computing entity to de-crypt these transmissions for fraudulent or other malicious purposes. The resulting reduction in fraud, not to mention waste and abuse, generated by GovGrid might save enough money to pay for the creation of the networks in the first place, and also free up additional capital to enhance and expand them.

Providing GovGrid and other grid-generated processing power to schools, universities, animation companies, public safety agencies, and massive multi-player environments to support the creation of interactive learning systems, cartoons, mapping of crimes, and synthetic visual and auditory environments would be just some of the ways grid computing power could offer benefits to individuals, groups, and governmental and commercial organizations.

California, home to Stanford University, the University of California system, Cal Tech and many other fine institutions of higher learning, a debtor state to the tune of ten or fifteen billion dollars, site of Silicon Valley and still capable of concerted innovation, ought to be the first to build a powerful state GovGrid. Generous Californians, with millions of computers in their offices, factories, and homes, ought to consider it a civic responsibility to donate their billions of unused microprocessor cycles to the state, to CaliGrid™, for the common good and the good of every citizen, student, worker, businessperson, artist and bureaucrat who will be able to make their own work and life easier through the new-found power of grid computing.