Where will the next generation of personal computers come from?

Probably Japan.

Where will the ideas for the next generation of personal computers come from?

Quite possibly, the new Media Laboratory at the Massachusetts Institute of Technology in Cambridge, Mass.

The laboratory has the potential to be the most intellectually exciting and provocative research domain in personal computing -- and Lord knows the personal-computer industry needs some new concepts and excitement.

The brainchild of MIT Prof. Nicholas Negroponte and MIT president emeritus Jerome B. Wiesner, the Media Lab is not the typical academic research ivory tower; nor is it a purely commercially oriented development organization. Researchers aren't looking at parallel processing or large-scale integrated circuits; they're examining how people can best use media technology.

That means there are research programs in speech recognition, holographic computer imagery, the design of man/machine interfaces, and how best to exploit the enormous data-storage capacity of compact laser discs.

In essence, the lab asks: How do you package the cutting-edge technologies in unusual and intriguing ways? How do you make the technology relevant to people who don't care about technology? How can you use technologies to add value to the print, computer, audio and video media?

(A digression: Some people love Apple Computer's Macintosh. They love the ease of interface; they love the mouse, they love the display. The machine "fits" their cognitive style.

As you may be aware, while the Macintosh came from Apple, the underlying concepts originally came from Xerox Corp.'s legendary Palo Alto Research Center.

PARC is one of the most fascinating technology stories of America's post-war history for a complete profile, see the October issue of Spectrum magazine . Xerox poured millions of dollars into PARC to attract the brightest minds and perform some of the most stunning computer research in the world. Xerox's PARC was a Xanadu of cutting-edge information technology. What was so unusual was that a profit-driven enterprise would invest so heavily in such fundamental and basic research.

Alas, Xerox was unable to commercialize many of its own creations successfully -- other companies and PARC alumni did instead. One cruel industry joke was that "PARC is a national resource for everyone but Xerox."

But what PARC did was vital. It became a key agent of technology development; it made technologies such as Macintosh and mice and laser printing possible and economical. End of digression.)

MIT's new media lab offers the same hope, and without the profit pressures that forced Xerox to slow its support for PARC. The lab has attracted some very impressive people -- including Alan Kay, the Xerox PARC alum who is the progenitor of the "dynabook," the concept of the personal computer that's as portable as a paperback but as versatile as a musical instrument, notepad and sketchbook.

The center has also attracted Marvin Minsky -- the eminence grise of artificial intelligence.

The center is stuffed with interesting projects as well. Newspeek is a comprehensive personal-computer system under development that blends a compact disc, high-resolution screens and artificial-intelligence software into a personal computer that becomes a personal newspaper editor. The computer analyzes all manner of news data flowing into it; arranges that data in order of importance; retrieves digitized photos from a compact disc archive to help illustrate the stories, lays out the stories and the photos in a newspaper format on the screen and updates it all as the latest news breaks. (Yes, but what about the advertisements?)

The point: The personal computer becomes a personalized, customized data-retrieval and organizing device. There's going to be a market for that.

Another project is Phoneslave. Don't think of Phoneslave as a personal computer -- think of it as a supersmart answering machine. Phoneslave answers the phone; uses voice synthesis to ask the caller to identify him or herself; checks its memory to see if there's a message that should be delivered to that caller, and then takes a message.

Phoneslave can build a database of calls, callers and messages while fitting them all into a software "calendar" module so that appointments and arrangements made over the phone can be stored in its memory for future reference. Similarly, Phoneslave can store voice "tickler memos" to remind you to bring up important subjects the next time you call a particular person. Those are just a couple of ideas that slowly, but surely, will become models for future personal-computing products. Presumably, MIT's rich academic and technical traditions will assure a flow of useful ideas from the Media Lab to the marketplace.

Can this lab give American companies an important edge in finding the future of media technology? Of course.

One sobering note, however: The largest corporate contributor to the new lab isn't IBM or Apple or Xerox or Digital Equipment Corp. -- it's Nippon Electric Corp., one of Japan's largest and most successful technology conglomerates.