A chorus of
digitally obsessed Delphic oracles swears the PC lingers on its
deathbed -- that the network and information appliances represent
the future of computing. In the past, this view hasn't exactly sat
well with Microsoft (MSFT)
and Intel (INTC).
But Intel, in addition to gobbling up companies that produce or
service data and voice communications networks, has recently broken
precedent by launching a research effort to explore this alternative
vision. "Our objective is to look for new uses and new users for our
microprocessors," explains Fred Pollack, director of Intel's
Microprocessor Research Laboratory (MRL). "So in this case we want
to look beyond PC computing."
A chorus of digitally obsessed
Delphic oracles swears the PC lingers on its deathbed.
|
The MRL, now 160 staffers strong, was established in 1995 to
explore longer range computing issues. This fledgling project has
been tentatively dubbed Smart Computing. But while it reflects the
general view of pervasive computing - where people interact with
machines on human terms and the computers seamlessly do the right
thing - the project's intent is less to pioneer new trends than to
make sure Intel's chips can handle them.
Therefore, Pollack has dispatched his forces to first probe only
one representative face of the vision: the home of the future. The
premise is this: Just as most households possess a central heating
unit, an increasing number will soon incude a central server,
complete with broadband Internet connections. A host of devices -
microphones, speakers, cameras, displays, appliances, televisions
and more - will be linked wirelessly to the server, removing many
inputs and outputs from the PC and creating a fundamentally new
human-computer interface.
Look Ma, No Hands
Intel's initial approach to this goal
is to create a "knowledge bus" for connecting functions such as
speech, video, vision recognition and 3D graphics to both a
distributed computer network and an information database that can
mine the Web for needed data. MRL wants to then test this software
backbone by prototyping three aspects of intelligent services:
financial, communications and entertainment.
Pollack explains these ideas by imagining himself returning home
from work. A front-door camera sends his image to the server, which
performs rapid-fire pattern recognition to identify which occupant
has arrived. The system then greets him by name and posts Intel's
closing stock price and the rest of Pollack's portfolio on the foyer
wall display. Important related news is also shown, including a
cheery item about Advanced Micro Devices' having problems with its
latest chip. Satisfied he's still solvent, Pollack enters the living
room, greets his wife and begins to unwind.
Just then, the computer informs him that his boss, Albert Yu, is
looking for him. All voice and e-mail communiqués are integrated, so
Pollack doesn't care which mode Yu used. He just says, "Play that
for me," and out comes Yu's message. Pollack then tells the system
to telephone his boss, so it places the call. After a brief
discussion about a key strategic hire, Pollack and his wife settle
down to dinner.
During the meal, they decide to go out to a movie. The system
knows not to select "The Matrix" (Pollack's wife hates science
fiction) but a romantic comedy rated highly by a trusted film
reviewer. The show must also be at a nearby cinema and start at a
convenient time. Such information is accessible via the Internet,
notes Pollack, but it would take many clicks to get him there. It's
far better to let the computer do it.
The key technological elements for making this scenario real -
the Internet, wireless and broadband communications, speech
recognition and natural language processing - are close at hand.
"The real challenge here, and it's a major challenge," Pollack
notes, "is the integration of these technologies." If that can be
achieved, the total available market for Intel's chips should expand
nicely. If that means the PC isn't quite the icon it has been, he
reasons, so be it.
