No matter how sharp, gorgeous and sexy its pictures may be, high-definition television is not going to be just another pretty interface.

In fact, for people to say that high-definition TV will be just like video but with ultra-sharp imagery is about as meaningful as some marketing geek back in 1948 claiming that these newfangled television sets are just like radios with pictures -- it's absolutely true, but it completely misses the point.

While it's easy to think of high-definition television as pretty pictures, it's smarter to think of it more as IBM TV, a potentially provocative blend of technologies and programming.

Technology is forcing a convergence of once separate industries: Computer displays are becoming more like television sets, which are becoming more like radar screens, which are becoming more like magnetic resonance scanners.

Just as compact discs capture sound and music in digital form, technology enables us to capture and transmit imagery in digital form. The question isn't just designing high-definition television sets -- it's creating high-resolution imaging systems. The challenge is coming up with an architecture that embraces this new convergence.

This is precisely why U.S. representatives from companies including International Business Machines Corp., Apple Computer Inc., Hewlett-Packard Co., Digital Equipment Corp., the cable television industry and broadcasters have quietly been meeting over the past few months. They want to become the Frank Lloyd Wrights and Le Corbusiers of this new digital-imaging architecture that will allow television pictures to flow seamlessly into computers -- where they could be digitally manipulated if desired -- and back again. Not incidentally, these people strongly believe that it's better for American companies to craft this architecture than the Europeans or the Japanese.

"For the first time," says Lee McKnight, a post-doctoral student at the Center for Technology, Policy and Industrial Development at Massachusetts Institute of Technology, "the broadcasters, cable industry, computer, telecommunications and satellite companies have all recognized that they have common interests in digital imaging that will shape the information economy of the future. They feel they will benefit by cooperatively developing these architectures."

A committee, which held a workshop this week in Washington under the auspices of the National Academy of Sciences and the Institute of Electrical and Electronics Engineers, is groping toward a consensus that would radically redefine the architectures of television sets and personal computers. (Do not adjust your sets; they are being adjusted for you.)

Today, the television is a sealed box that can receive a broadcast transmission, a cable link and a VCR hook-up. But why not a TV set designed along the "open architecture" principles of the IBM personal computer? If you want to turn the TV into a computer, drop in a special board. Do you want your set to receive a high-definition TV transmission? Plug in the hi-def attachment. Want to build in freeze-frame, replay and editing capabilities into your television-computer? Just make the necessary attachments. In other words, the TV isn't just a TV anymore; it's part of an architecture that gives you as much power and versatility as you'd like.

Conversely, take your IBM PC or Apple Macintosh and plug in the board that turns it into a TV so that you can make multimedia presentations. The point is that once you have a defined architecture -- and the technical standards that go along with it -- you have created a new infrastructure for innovation.

This is critical because -- to be blunt -- we don't have a consumer electronics industry in this country anymore. Existing television standards (which were slapped together rather than built with any notion of systems architecture in mind) basically reinforce the notion of television sets as sealed boxes.

By breaking that model, U.S. companies have a better chance to carve out niches in high-resolution displays, subsystems such as image enhancing systems and remote control devices, and, of course, programs that run on both the computer and television sides of this next-generation device.

"The architecture must also accommodate whatever exists today and permit graceful transition from today's 'tower of babble' to the seamless inter-operatability that we need for the future," says Alan K. McAdams, workshop chairman and visiting fellow at the Economic Strategy Institute in Washington.

In many respects, this architectural approach is a necessary first step that U.S. companies have to take in order to hope to compete in this newly convergent marketplace. Certainly the Toshibas, NECs, Hitachis and N.V. Philipses of the world are already well-positioned to participate in these markets.

However, they have not yet demonstrated the same grasp of systems architecture that successful American companies have historically had. America's unique competence in this domain gives our companies a fighting chance to be a real leader of these emerging hybrid technologies.

Michael Schrage is a columnist for the Los Angeles Times.