Today's multimedia industry is being formed by the convergence of three separate, vertical industries: telephone, television and personal computing.

Each industry's product and content were originally as distinct as those of the airline and pipeline industries. But the rapid increase in computers' processing speed, the dramatic decline in the cost of processing and memory capacity, and advances in transmission technology have obliterated the traditional boundaries. As a result, information-age conglomerates of as yet unclear shape and field of activity are emerging.Digitization techniques have rendered analog signals, transmitted by telephone and television, indistinguishable from digital signals transmitted by personal computers. The result is that the same infrastructure accommodates manipulation and transmission - if not yet the input and display - of voice, video and data. Input and display are converging as video-conferencing combines voice and video in one transmission, while voice and handwriting- recognition capabilities combine, albeit somewhat imperfectly, voice and data.

The essence of the revolution is that content-specific distribution is being transformed into content-independent distribution provided by a common infrastructure. Thus the multimedia industry is emerging not as a set of three vertical businesses, but as five horizontal activities that together deliver content to consumers. They include content, packaging, transmission, manipulation and terminals.

We believe that the supply of content will, over time, remain closest to its current form. While the multimedia revolution will overthrow existing distribution advantages, the key factors in success will still be those that lead to the production of the highest-quality content at reasonable prices.

The Disney-Capital Cities merger and Time Warner's merger with Turner Broadcasting vividly illustrate the packager's and the broadcaster's need for assured access to high-quality content. These mergers will not be the last. In fact, they will accelerate the search for content providers by other industry players. Premium content providers will be in the driver's seat and should agree to participate only if offered substantial rewards above independent operation.

There is today no dominant packager from any of the three vertical industries. As a result, computer-software companies, cable and television programmers, telephone and online networks and myriad other entrants, possibly including electric utilities, will end up competing for control. Indeed, they are already encroaching on each other's territory (for instance, Sega's move into cable and Bertelsmann's into online services.)

To be profitable, packagers must develop concepts that override the reputation (or brand loyalty) of content providers and that replace the direct customer contact enjoyed by transmission companies. In the short term, successful packagers will be allies of high-quality content providers and transmission companies.

The transmission infrastructure will remain a shared function - a public network. This is not to say that there will be only one monopolistic network, but rather that each digitized network will be able to carry any content. These networks will have inherent economies of scale and unlimited incremental capacity.

The two most important questions regarding the future of transmission, then, are how many local networks, and of what type, will emerge, and whether local and long-distance service will remain separate.

Transmission, meanwhile, is fast becoming a commodity, with prices dropping dramatically.

As to the manipulation infrastructure, a real distinction has arisen between processing and storage hardware and the intelligence of the software that governs the manipulation of the data. In the multimedia industry, just two main kinds of software will be needed. The first, which we might call ''information-superhighway software," will enable that highway to function as much more than a simple conduit for data. This hybrid, combining elements of both manipulation and transmission, is a battlefield on which those industries are vying for position. The second category, "traditional manipulation software," includes operating systems and stand-alone applications.

The terminals market is almost evenly divided among the telephone, television, and personal-computer industries. That suggests that the universal terminals market (for instance, the set-top box) is up for grabs. Market fragmentation is likely as terminals serving specific segments will proliferate.

The most important difference driving segmentation is the wireless (portable) vs. wireline (fixed location) dichotomy. Some terminals will be multifunctional, but differentiated consumer demands will support large numbers of specialized competitors.

Content providers are the focus of the most intense and growing attention, as packagers and transmission companies try to lock in high-quality content to differentiate their consumer offerings.

Packagers should be involved in alliances with content providers to differentiate product offerings, and with transmission companies to incorporate technological knowledge of network capabilities into their creative activities.

Transmission companies should join forces. AT&T-McCaw and US West-Time Warner are the obvious prototypes. Wireless companies need to build nationwide networks, and cable companies need to develop national distribution scale and switching capabilities through alliances with telephony companies.

Manipulation and terminal providers are attractive partners for transmission providers and packagers seeking assistance in developing standards. Manipulation-transmission alliances, such as those that Netscape has already formed with a variety of companies, have yet to arise in great number, but we expect to see many.

Industry convergence is inspiring a great deal of radical reorganization in many companies, as well as consolidation within and between industries.

Although the uncertainty that remains is tremendous, so are the rewards for those with staying power.