The Internet is not only the data network as we know it today, which provides principally e-mail services and access to sites on the World Wide Web, but also the socio-economic phenomenon unleashed by the ongoing networking of all PC and data-processing units.
This technological revolution comes in the wake of the decentralization of mainframes and of the development of personal computers. Hardware and software are produced with extreme economies of scale. Software or microprocessors are very costly in terms of research and development, whereas their marginal cost of production is very low indeed. The need for initial expenditure to be offset by mass production has brought about widespread distribution of personal computers, firstly in companies and then in the home. Thus, today’s installed computation capacity is under-utilized - most personal computers do not operate for more than a few hours a day at best. This is in stark contrast with the mainframes which paid their way by operating day and night, just like blast furnaces.
Similarly, whereas data processing was originally a trade for specialists, personal computers have meant that software has become more user-friendly and interfaces more ergonomic (icons, mouse, etc.). A significant proportion of the labor force has grown accustomed to using microcomputers in the workplace and a large number of amateur computer users are now capable of taking part, if not in program writing, at least in testing beta versions.
The Internet is now benefiting from this pool of available experts and computation capacity. It is also able to draw on rapid technical progress which will enable it progressively to extend the available bandwidth without having to increase the access price or to use usage-based tariff systems. That is to say:
transmission costs are coming down very considerably, which means that distance is less and less important in overall costs (at least for a constant bandwidth);
switching costs are also coming down. Moreover, in packet technology, such costs are less dependent on usage and more so on access;
local distribution (over a copper loop, possibly improved by the use of ADSL) and packet switching allow the end user to be connected 24 hours a day: computers can remain permanently linked to the network, traffic costs depending only on the volume of information actually exchanged;
the reduction in costs of microprocessors and the progress achieved in the field of software make it possible to imagine that semantic processing of information will soon become available (leading to sophisticated search engines and archives for texts and still or moving images).
The development of the Internet also goes hand in hand with a new attitude towards innovation. The industrial revolution established an ideology: economize on physical work so that no man should have to toil as a slave. Computing extended this worthy aim to repetitive, white collars tasks, i.e. office work in its simplest and most repetitive form.
Networking has established quite a different ideology which has long been presented and discussed on the Internet15: ensuring that everybody can benefit from the free flow of information permitted by the development of digital networks. This means allowing the free exchange of information, independent of any censorship, supervision or protection. Beyond the debate on sexually explicit or potentially offensive materials, this ideology questions the very principles of intellectual property rights (copyright, protection of software, etc.) and, ultimately, the property rights of companies on their internal information.
Digital networks allow markets to operate with much greater fluidity in that:
information on products is easier to gather. Consequently, the consumer is able to compare a larger number of available items and can also gather information very rapidly about the launching of new products, which makes markets more dynamic and more reactive;
for a given product, prices can be compared very easily. Search engines can already provide suppliers’ addresses, classified according to what they offer in terms of price and sales conditions. They will be able to extend this assistance towards more qualitative comparisons similar to that of a hotel guide containing subjective information on a town’s restaurants and accommodation;
it will be possible for new types of trade to be developed, for example sales by auction for second-hand goods or products not commonly found on the market;
more generally, information on the market and the search for better purchasing conditions will progressively be performed by software (negotiating agents) responsible for carrying out transactions. In this case, it is possible to imagine that complex contracts will be negotiated, taking all the factors of the vendor/purchaser relationship into account (credit, commitment to after-sales service, maintenance, hot-line service, long-term commitment between purchaser and supplier, etc.).
On the other hand, networks may not be sufficient to create pure and perfect markets. Trade will undoubtedly be more fluid, but companies will be able to utilize digital networks to define highly focused product differentiation and to carry out one-to-one marketing.16
consolidate temporary advantages into long-term dominant positions. On a fluid market, a minor discrepancy in price or service quality is enough for demand to concentrate on the best vendor. A slight or temporary advantage will then be sufficient to create a permanent monopoly (a sort of “winner-take-all" effect).