archive:news:bcwired
WHAT DOES A NOBEL PRIZE FOR RADIO ASTRONOMY HAVE TO DO WITH YOUR TELEPHONE?
It's been a decade since the breakup of AT&T. Has the spirit passed
out of its Bell Labs, as some charge? Or is it still the preeminent
technology lab in the US, only "more nimble, more intelligent"?
By Richard Rapaport
_________________________________________________________________
Edward Eckert walks through a storage area behind a yellow
corrugated-metal warehouse set off a wooded road in piney Warren, New
Jersey. He presses ahead, into a cold storage area filled by a
room-sized mobile file cabinet, set on tracks. This is the archive of
Bell Laboratories, the near-mythological research arm of American
Telephone & Telegraph. Spinning a lever that opens the space between
two of the 7-foot-high files, Eckert pulls out an inexpensive black
Naugahyde case, the kind you can buy at Woolworth's for 15 bucks. He
unzips the case and fishes out an ancient, fraying 3-inch-by-5-inch
notebook in purple alligator-patterned leather. Opening it, he
delicately pages through.
The notebook's yellow-gray pages are lined with faded green ink and
begin with a penciled date, February 1876. Only the first 20 pages of
the notebook are used, filled with the minutiae of a young Boston
laboratory assistant named Thomas Watson, who made his recordings
during the winter of the American Centennial. There are a few simple
sketches of electrical devices - switches and the like. There is a
list of expenses that the frugal 22-year-old incurred during that
winter: "tooth powder - 35," "ice - 10," and even an entry for
"drawers - 1."
Cradling the notebook, Eckert turns to the fourth page. At the top is
the date - March 10, 1876. Below, in tiny script, are seven words that
are among the most momentous in the annals of science: "Mr. Watson
come here I want you."
Such were the words of Watson's employer, a 29-year-old Scottish-born
inventor named Alexander Graham Bell. They were spoken from one room
to another in Bell's laboratory at 5 Exeter Place. The tale is a
staple of scientific lore: Bell really did need Watson, having just
spilled acid on himself. More importantly, his call for help was the
very first electrically transmitted message spoken over Bell's
instrument, soon and forever after known as the telephone.
Other entries on the fourth page of Watson's notebook record more of
that day's prosaic, yet epochal communication. "How do you do," Watson
chronicles. "God save the Queen and several other articulated
sentences," is the final, triumphant entry on the page of the aged
notebook, only one of many rarely viewed prizes tucked away in this
nondescript warehouse in a clearing carved out of the New Jersey
woods.
The artifacts collected here catalog the output of Mr. Watson and his
descendants, the thousands of scientists who, throughout most of this
century, gave Bell Labs nearly as much of a monopoly on scientific
innovation as its parent, AT&T, had on international communications.
The caged and locked holds in the Warren warehouse smack palpably of
the Raiders of the Lost Ark, containing as they do prototypes that do
nothing less than define the course of our technological century.
There, in boxes, is the world's first carbon dioxide laser; on a dusty
table sits the original quartz clock; on a shelf rests the telephone
on which the first transatlantic call was made; scattered about are a
pilot's helmet used for the first ground-to-air radio transmission, an
early loudspeaker used at President Warren G. Harding's inauguration,
the world's first solar battery, and the original artificial larynx.
The archive is the resting place for quaint technological failures as
well - items like the sadly aged 1954 prototype Picture Phone. Then
there are the objects whose ubiquity underscores their importance: the
coffin telephone booth, the black 500 Series Western Electric
telephone (it once graced every office and home in America), and the
green, orange, blue, and white versions of the mod, quintessentially
'60s Princess phone.
But here in Mother Bell's basement, the most awesome artifacts of all
are the more than 100,000 scientific notebooks - tucked away in row
after two-story, pre-fab, metal-frame row - that hold the notations of
Bell Labs scientists, modern-day Leonardos whose theories,
discoveries, and inventions have immeasurably altered humanity.
Since its formal incorporation in 1925, Bell Labs's scientific
fraternity represents a Who's Who of international research. It
includes seven Nobel Prize winners: William Shockley, Walter Brattain,
and John Bardeen, inventors of the transistor; Clinton Davisson, who
demonstrated the wave nature of matter; Arno Penzias and Robert
Wilson, whose work in radio astronomy confirmed the big-bang theory;
and Philip Anderson, for his work on the deep atomic structures of
metals.
Notes of the basic research of these Nobel laureates and other
scientific giants make the Bell Labs archive a veritable pantheon of
technological achievement. There is the work of Claude Shannon, whose
seminal information theory provided the framework for computer
programming; the research of William Pfann, whose "zone refining"
process made possible the mass production of semiconductors; the
formulations of Alfred Cho, whose molecular beam epitaxy allowed
microprocessors to shrink to undreamed-of size and complexity. There
are the scribblings of Ken Thompson, who, along with Dennis Ritchie,
developed Unix, the first cross-platform computer operating system;
and the notebooks of Bjarne Stroustrup, father of the key programming
language C++.
The archive, in total, houses the intellectual foundations of more
than 25,000 patents, nearly one for each day of Bell Labs's existence.
It is an outpouring of scientific innovation that -- in its
breathtaking variety and willingness to push beyond the temporal world
of telephony - has more than lived up to Alexander Graham Bell's
refreshing exhortation: "Leave the beaten path and dive into the
woods."
Bell's dictum is carved into the pedestal of his bust in the lobby of
the sprawling Murray Hill, New Jersey, compound, one of a cluster of
campuses where many of Bell Labs's 25,000 employees work - all within
an hour of the institution's original New York home. But if Bell's
words are cast in stone, much has changed since the time when the
steely monopoly that controlled - nay, owned - virtually every phone,
line, pole, switching station, and PBX in America, showered its
largesse upon Bell Labs, making it, according to one executive, "the
world's best university." If Bell Labs's university lacks students, it
comes complete with in-house physics, art, radio-astronomy,
astrophysics, and its own economics department.
The golden age of pure research at Bell Labs began in the mid-'30s,
when the search for a solid-state device to replace the vacuum tube
began. It was a spirited period when, one former scientist recalls,
"people rode unicycles in the halls and invented mind-reading
machines." It lasted into the late '50s, when Sputnik-phobia drafted
Bell Labs into the Cold War, even as it remained a bastion of reason
and safety against the loyalty oaths that were requisite in many
American institutions for years during the frenzy of McCarthyism.
Though staff sizes and budgets remained generous afterward, subtle
changes in mission and status came along with the anti-establishment,
anti-scientific Luddism of the late '60s. Regulatory hearings into
Bell system rate-making in the '70s further diminished its hegemony.
Then came the 1982 federal court consent decree that broke up the Bell
system, awarding Bell Labs to AT&T, while setting up a parallel
organization, Bellcore, as the research wing of the so-called "Baby
Bells." The transformation accelerated during subsequent internal
reformations at AT&T, as the company slowly moved away from its
research orientation and sought to evolve from ponderous scientific
sovereign into successful business and technology competitor.
In his searching 1984 book on Bell Labs, Three Degrees Above Zero,
Jeremy Bernstein posed a troubling scenario about the effect AT&T's
breakup would have on the scientific crown jewel that was Bell Labs.
"Clearly with the divestiture," Bernstein wrote, "Bell Laboratories is
at a watershed. If all goes well it can continue its great tradition
of basic and applied research, and if things do not go well it runs
the risk of becoming just another large, conventional industrial
laboratory."
Has the past decade written the d?nouement in the drama of Bell
Laboratories's event-horizon research? Other phone companies, like MCI
and Sprint, have become profitable and competitive - while providing
little or none of their own basic research - and have cut into the
AT&T profit margins that fund Bell Labs. Bell Labs nevertheless
remains a scientific hothouse, with a US$3 billion total budget - $2.7
billion for development and $300 million for research. It is the
richest and largest private research lab in the world, still leading
in fields as diverse as photonics, fiber optics, HDTV, artificial
intelligence, cellular telephony, digital radio, and computer
software.
Still, Bernstein's decade-old rumination resonates with some veteran
Bell Labs scientists and executives, who are steeped in the Labs's
research tradition and watch each change with some trepidation. Are
things going grandly? In the words of current vice president for
research and Nobel laureate Arno Penzias, is "the company more nimble,
the academy more intelligent"?
Or has Bell Labs conceded its preeminence in basic science, opting
instead for the kinds of business-driven technological innovations
that can come to term in months instead of years or decades; has the
institution opted, in other words, against the lightness of the
scientific and toward the gravitational pull of the commercial?
Bob Lucky, Arno Penzias's counterpart at Bellcore, believes the
latter. Lucky, who served at Bell Labs for 31 years, is one of a
number of "old-timers" who longs for "the golden years," believing
that "the spirit has gone out of the place, a spirit whose passing
people mourn."
Lucky is quick to point out that many of the changes at Bell Labs were
necessitated by AT&T's own deregulated, diminished status. And he
lauds the current AT&T chair, Bob Allen, for fighting hard for the
Labs US funding and staffing. But with a new president, Daniel
Stanzione, just beginning his administration, Lucky and others wonder
if Bell Labs is not necessarily in for a downsizing or a restructuring
that will see an increasing number of scientists attached directly to
business units rather than remaining in independent research groups.
For his part, Stanzione tries to answer Bernstein's question by
talking reassuringly about maintaining Bell Labs's historic commitment
to scientific independence. "We have continued the tradition of both
basic and applied research," he contends, "with the percentage of our
budget still going to meet the letter of what (Bernstein) suggests."
Bell Labs's eighth president, Stanzione was most recently president of AT&T Network Systems's Global Public Networks. Despite 17 years as a computer scientist at Bell Labs, Stanzione's concern for the spirit of scientific independence jibes most with the newer, business-?ber-alles model of the place: "The measure of our success," he says, "will come as we move forward and see whether the technology that comes out of basic research will prove advantageous to AT&T." His point: "To simply provide technology in broad ways and infuse the industry - that's not the way our industry works today."
Stanzione's pledge to maintain integrity notwithstanding, Bob Lucky cannot help but fondly recall the old Bell Labs way as exemplified by Bill Baker, the legendary president of the institution in the early '70s. Party to a nuisance suit, Baker was being grilled about his responsibilities during a deposition. As president, he was asked, wasn't it his job to manage people? Baker replied with what he, at least, felt should be obvious. "People know what to do," he told his inquisitor, referring to Alexander Graham Bell's prescription to seek out the kind of men and women who, more than anything else, were unafraid to "leave the beaten path and dive into the woods."
Where to go for ideas
"I find it hard to believe that people who are idea-driven are not entrepreneurs today," Waring Partridge, AT&T's WASPily articulate vice president for multimedia strategy, suggests over mid-morning coffee. We are sitting in an empty cafeteria within the central atrium of the sprawling, brown-brick cluster that is Bell Laboratories's Murray Hill campus. Partridge, whose group brings to market AT&T consumer multimedia services, has come over from his Basking Ridge offices to tap some of the rich Bell Labs brainpower, as well as to talk with me about what he sees as the "new tradition" of AT&T business/science dual citizenship - a trend that has seen many once-independent Bell Labs researchers go to work directly for AT&T business units. Partridge is a new phenomenon at AT&T. He is an entrepreneur, a former McKinsey management consultant and the founder of small telecommunications, paging, and cable companies. "Instead of picking up my chips and playing golf," he says, he decided, at the age of 49, to sign on to one of the world's least entrepreneurial mega-corporations. "In the '80s, I said I'd never work for AT&T," Partridge recalls about a company that, to his mind, is still "commercially underdeveloped." But when Partridge, who looks and sounds a bit like a techno-George Plimpton, finally did take a look at the new, deregulated AT&T, he liked what he saw: for starters, a company spending $3 billion a year on software alone, with the resources and staying power to force what he believes will be the technology passage to the future: the convergence of telecommunications, broadcasting, and computing. Like many of his AT&T colleagues, when Partridge looks at such soon-upon-us wonders as interactive TV, he doesn't see television, he sees telephone. During his four years with AT&T, Partridge has become a Bell Labs partisan. It is, he says, the kind of place "I go for ideas." But Partridge is critical of Bell Labs's history of noncompetition, when "people worked so far in the future they could spend their entire lives and see their technology in the market only after they retired." His conception of the new Bell Labs mirrors that of many of AT&T's and Bell Labs's managers.
"It's a question of what is needed, more than what is possible," Partridge insists. He is unapologetic about strengthening the arrangement under which, as he suggests, research staff "reports to the business units that give them money." And that is why, finally, Partridge cannot help but latch onto the notion of Bell Labs as an entrepreneurial center, a place where, more and more, scientists and researchers are allied and aligned with the business units.
He expresses his hopes for this notion by citing examples of Bell Labs scientists who have taken initiative, saying to their marketing counterparts, "We have the technology, what are you going to do about it?" And he praises Labs Vice President of Research Arno Penzias for his support, telling the story about Penzias's beyond-the-call-of-duty endorsement of AT&T's new PC-based telephone videoconferencing system, Vistium.
Prior to the announcement, in the summer of 1994, it was Penzias who brought a couple of Vistium terminals over to AT&T's Basking Ridge complex and set them up. Then Penzias, the man who won a Nobel for finding proof of the big-bang theory, proceeded to walk up and down the halls, banging on executives' doors, bringing them into the conference room to give Vistium a try. Throughout the conversation, Partridge returns to one point: the successful technology company of the future will likely emerge, not from the computer industry, with its vested interest in the microprocessor box, but from the telecommunications industry, with its tradition of networking and interoperability. He also believes that AT&T's long-standing "plug and play" credo, which eschews instruction booklets, will be the successful future technological research and development paradigm. "Most things that succeed," he suggests drolly, "don't require retraining 250 million people." The great intelligence of the network Waring Partridge's point is also Eric Sumner's. Sumner is demonstrating the Sage Project - recently rechristened the more commercial "AT&T TV Information Center" - in a comfortable, gray- and cranberry-accented loung area up the hill from the main Murray Hill complex, inside the sparkling new Consumer Lab. "It's the AT&T way: no manual," says the product development vice president for intelligent systems, as he points a handheld control at a virtual remote flickering on the screen of a huge Sony television.
The TV Information Center is a new AT&T service introduced in January 1995. It is designed to provide onscreen messaging; rudimentary weather; stock, news, or traffic information; but not, Sumner hastens to add, visual phone service. Sumner and his colleagues are charged with developing and marketing the Information Center and the next generation of telephones, a seminal AT&T ritual that comes around only once every 30 years or so. He traces the evolution of the phone, beginning at the end of the last century with the crank devices, evolving to rotary telephones in the '30s, touch tones in the '60s, and, Sumner hopes, to other Information Center-like screen-display products for the '90s and beyond.
Unlike earlier eras, though, this time AT&T is no longer a monopoly, says Sumner: it "can't dictate." And, heeding the early '60s disaster of the Picture Phone (which Sumner suggests never took off because "video phones required new behavior"), Consumer Lab researchers are careful to use technology already familiar to the average American consumer - like the ATM, the television, and the remote control.
Indeed, little would fluster the average PC user in Sumner's demo. With the push of a button, the TV Information Center can record, store, and then visually render telephone messages or faxes; it can be programmed to automatically dial up and store voice and text services; it can give an instant readout on all the day's relevant stats - before you've even brewed your morning coffee, something it cannot do.
The Information Center's simplicity - you plug a black box into a phone jack and then into your television - is hardly accidental. Technological "ease of use," to borrow the jargon, is a major, perhaps the major, new Bell Labs paradigm. Retiring Labs president John Mayo compares its importance to "improving the efficiency of the vacuum tube in the '40s." Sumner once worked under Mayo at the Labs, but transferred to the AT&T consumer side when, he says, "a cool new head of consumer products brought me over here." But he didn't exactly leave Bell Labs behind: he forages there frequently, "looking for what needs to be built, and then wandering the halls looking for someone to build it." Sumner has found a collaborator in Thaddeus Kowalski, chief architect for products, into whose PC- and circuit-board-strewn lab we wander after the TV Information Center demonstration. "This shop allows more practical people to have an effect," the stocky, intense Kowalski says about products that have come out of his laboratory, as he catalogs existing pieces of Bell Labs technology that went into the Information Center. "We borrowed the graphics, we lifted the computer codes wholesale, and we already have the best transmission and file-compression techniques," he says. He winds up echoing the typical Bell Labs attitude: "People in the research area had all sorts of technology, and they were eager to get it out." Kowalski and Sumner are especially jazzed about another Sage Project product that will allow the software in AT&T devices to be automatically upgraded over telephone lines by modem. "We can download new codes to older machines," Sumner points out. "It's nice for consumers that Information Center boxes are less likely to become obsolete."
This process is a critical component for what Sumner sees as the
network-driven "software delivery model of the future." He believes it
will allow millions of customers to "hook into the intelligence of the
network."
Plan 9 From Outer Space
Murray Hill is a maze of industrial structures rivaling the Pentagon in sheer potential for misdirection. If you are able to find Corridor C, on the fifth floor of Wing 2, you will discover a commons room in which computer scientists have long congregated for informal parleys, the custom for disseminating information at an institution renowned for openness and collaboration. For years, this gathering spot was known as the Unix Room: chats held here in the '60s by programmers including Ken Thompson and Dennis Ritchie spawned the Unix operating system, the first software that allowed different kinds of computers to talk to each other.
Today, the informal name of the area has changed. It is now know as "Plan 9 Land," a tribute to the new operating system that Dennis Ritchie's computer techniques research department designed and named after Ed Wood Jr.'s legendarily awful science fiction epic, Plan 9 From Outer Space.
Plan 9, the system, not the movie, is a descendant of Unix. It is a
shell program that allows different and differently programmed devices
the freedom to network and process distributively in the simplest and
most economical way ever.
Allowing transparent distributed processing over a network, Plan 9
already drives such products as the TV Information Center and AT&T's
World Wide Web 800-number directory. It will, according to Ritchie,
give users the freedom to work simultaneously with different
processing activities taking place in varying locations, but, to the
user, still be part of a unified activity. This is handy, he explains,
because "sometimes it's better for the data to be near the CPU and
sometimes it's better - in graphics, for example - to be near the
user." With Plan 9, the user is unconscious of where the processing is
taking place -- he or she knows only that a complex job is getting
done.
I ask Dennis Ritchie about the operating system's curious name.
Sitting in his tiny, cluttered office in Plan 9 Land, around the
corner from the former Unix Room, he answers by rubbing his chin and
repeating my question out loud. In his 50s, impish, bearded, and
dressed down, Ritchie scrunches up his face as he thinks, rattling off
other whimsical movie names his group has given their creations: titles
like 8 12, Brazil, Rio.
Ritchie shakes his head and launches into what he calls "the very complicated history of Unix," a system as famous for its revolutionary notion of separating hardware from software as it is for the fact that AT&T lost untold millions unsuccessfully trying to license and sell it.
Not that Unix's failure to make money ever dented Ritchie's reputation at Bell Labs: in 1983, he won the ACM Turing Award, the same year he became a Bell Labs Fellow; since then, he has watched his group become a cornerstone for what is today Bell Labs's most prestigious department, computing science research. Part of the reason for the department's good standing at the institution is that so much of AT&T's business is now tied to software development and implementation, even a small tweaking of a telephone-switching or billing system can result in mega-savings. "Thousands of people are writing code: anything we can do to improve their product means we have justified our existence," Ritchie notes with a satisfied air.
Ten years after sending Unix out to its glorious and unprofitable fate, Ritchie admits, he and his group were getting bored. "People wanted to do something new, not necessarily with a product orientation," he says. "We thought the current versions of Unix and other operating systems were too big. It was an engineering issue to design more powerful features with a simpler code." The result, underway in 1988, was Plan 9. Ritchie is modest when he compares his hopes for the new operating system with the enormous impact of Unix. "The main effect of Unix, the portability of an operating system, can be done only once," he says. But Dennis Ritchie's unassuming description - he's trying hard not to scare the business types who are still calculating Unix's losses - belies the great excitement in Bell Labs about the potential for Plan 9. Beginning this year, Bell Labs will be offering Plan 9 on CD-ROM, with source codes and manuals available to developers at a modest price. The idea, Ritchie suggests, is twofold. First, he says, "it is a way to get it out and make it visible." And second, "by making it more available outside, it will make it more credible inside." Plan 9's importance may be that it is an operating system designed from the start to allow computers and communications devices to work together in tandem. This, of course, advances AT&T's prime directive: complete the interoperability for more and more of the highly profitable "fat minutes" that come from the processing and transfer of broadband data, naturally, across the AT&T network.
Firewalls and Bell fellows
It's noon at the Bell Labs. A score of scientists are gathered intently around chess boards set up on a balcony overlooking the lunchroom, which is fast filling with a bustling, heterogeneous mix of researchers and executives. One game in particular seems to be gaining steam: intent players are making aggressive, knowing moves and finishing off by slamming their palms down to stop the chess clock. Bill Cheswick, a 43-year-old programmer in the research computation department, stops for a moment on his way into lunch to contemplate the match.
Cheswick, co-author of the recently published Firewalls and Internet Security: Repelling the Wily Hacker, is an expert on computer-security gateways. He is equally well versed in the normally amicable chess game between pure research and industrial development that has played out between managers and scientists for nearly as long as there has been a Bell Labs.
Over a minimalist lunch of cottage cheese, bread, and butter, Cheswick describes the inside-out world at Bell Labs where the MO is "we tell management what we're going to do." Or, as Cheswick restates it, "Management presents opportunity; you get to choose your work." This attitude is a holdover from the old Bell Labs culture, which Jeremy Bernstein identified this way: "Either you do something very useful or you do something very beautiful." But perhaps the era of downsizing has had a discernible effect at the Labs. "It's possible," Cheswick says, "if you're bright enough, to spend the year fishing on a boat, then write a three-page paper, and get your salary. "But it's risky," he adds with thoughtful, practiced timing. These days, someone like Bill Cheswick is happy to avoid those risks, so delighted is he with his job and the encouragement Bell Labs gave him to write his book "on company time." Over a dessert of ice cream (a soda fountain's worth of flavors and toppings is kept in a nearby freezer), Cheswick spends a moment explaining the seemingly salubrious arrangement in which "you keep the royalties, and the company keeps the copyright." "It's a good deal," he notes. As are other perks, including his ability to acquire equipment. Cheswick, with only a BA, has the authorization to request "any kind of computer or software I want with less review than I have with my wife when we go shopping." He logs into a new $20,000 3430 Dual CPU NCR server to create extremely fast Internet security gateways. "I can cut as much rope as I want," he smiles, "and hang myself with it." Aside from the potential for scientific self-strangulation, Cheswick does have a few complaints with the current system. "We are part of the overhead," he explains. "Therefore, you get bean counters telling you, 'I'm paying a percentage of my profits for this, what am I getting?'" But in general, the prestige, salary, informality - today Cheswick is wearing a T-shirt, jeans, and sandals - and interesting assignments at Bell Labs make for an unusually contented and motivated work force. This includes those still huddled over chess as Cheswick excuses himself and walks out. As he passes, another hard-fought game is ending. "Nothing I can do," one of the players shrugs to his audience as he concedes the game. "I didn't use my bishops very well."
Pioneer One
Down one, then another, and yet another long tiled corridor, past nitrogen tanks, emergency showers, and innumerable laboratory doors, is the Elephant House, so called because of its 30-foot-high ceilings and exposed pipes. The place at Murray Hill where Bell Labs technicians still get their hands dirty, the Elephant House seems more like a maze for white mice, so chaotically and indifferently is its floor space divided.
Inside one of these chock-full-of-technology warrens is the Light Wave Lab, where Frank DiMarcello, technical manager for the optical-fiber research department, is finding ways to improve the manufacture of optical communications fibers. These filaments of highly reflective, light-bearing glass are rapidly replacing copper wire as the neurons of the AT&T network, giving it the increased bandwidth that is the key to the wired future.
DiMarcello works amid the sound of arcing electricity and the odor of ozone. The centerpiece of his domain is a 29-foot fiber-pulling tower topped with a furnace: rods of incredibly pure silica are heated and then drawn down through the tower, untouched by human hands, into a continuous 10-kilometer fiber that is 125 microns thick.
It is through these fibers that lasers can shoot highly tuned light
waves carrying an unprecedented and ever-increasing amount of
information. Information will travel in light pulses, each pulse
moving as much as 40 billion bits a second. This is the equivalent of
2.5 million simultaneous telephone conversations over a single strand.
DiMarcello fingers a length of optical fiber: it looks like
monofilament fishing line and can be tied into knots and still
transmit bursts of photons, carrying a thick bandwidth of information,
audio, text, or video. Millions of miles of such fiber-optic line will
be required to fiber the nation. AT&T produces all of its fiber at
just one specialized facility in Georgia. The race to develop faster
lasers to pump more data over fiber-optic lines is matched by a
separate race to speed up the fiber-production process.
"We're using larger tubes to try to make 100 kilometers of fiber with
each," says the thin, intense DiMarcello over the hum, as he picks up
a 4-inch-thick-by-3-foot-long glass pre-form, looking very much like an
icicle, that will be melted at the top of the tower.
Then he walks me through a door into another room in the Elephant House, to a rocket-shaped 70-foot tower. When it is put into operation, it will be capable of producing 15 meters of optical fiber a second, far more than the 29-foot Pioneer One we just left behind. "We're trying to increase the draw-speed," DiMarcello says, gesturing at the huge cylindrical tower. "The big tower gives the polymer extra time to cool."
Watermarks and "The Wasteland"
Dave Kristol is demonstrating new Bell Labs technology that creates an electronic "watermark," consisting of small shifts in the spacing of text that will allow publishers to identify and trace individual copies of works that have been printed from computer files. If successful, the watermark will be a viable mechanism allowing payment to be extracted for printing specific pieces of text off a computer network; it will prevent illegal copying and allow for the collection of royalties.
"If you hand out copies to 1,000 of your best friends, we can ask you to pay," says the Young Turk in the distributed systems research department. Then Kristol laughs and describes another feature his group is working on. "We call it an anonymous credit card," he says about a system in which "a person online can pay, but the seller doesn't know to whom his material is going."
As he speaks, Kristol continues working his way through a demonstration of Bell Labs's SEPTEMBER Project. The nightmare of every online freedom fighter, Bell Labs's "Secure Electronic Publishing Trial" project offers the possibility of bringing the electronic network into the realm of laissez faire economics and making "pay per piece" electronic publishing possible. If electronic copywriting and producing online credit cards is his official raison d'?tre, Kristol is more than happy to move on to demonstrate his real love - audio and video over the World Wide Web, into which AT&T has jumped with a vengeance. On his workstation, Kristol calls up a short video of AT&T chair Bob Allen giving a speech. It is being sent from a server at another Bell Labs campus and is part of what is known as "Nemesis," a network-friendly service that allows video and audio to be sent and viewed straight off a server as needed, rather than forcing a user to download an entire file first. Kristol is equally proud of the Bell Labs's version of Internet talk radio and the in-house "jukebox," a server on which various recordings are stored. He plays a portion of the radio series Hell's Bells: A Radio History of the Telephone, before getting to the p??ce de r?sistance - over a speaker hooked up to Kristol's workstation comes a ghostly version of T. S. Eliot's signal poem, "The Wasteland." The solemn, late autumn sky outside Kristol's office window seems to close in on us as Eliot's reedy voice intones, "Here is no water, but only rock."
While it is not quite "Mr. Watson come here I want you," the analogy is clear and powerful.
Not as OK as it used to be
Through the floor-to-ceiling windows in Arno Penzias's office, the setting sun glows fiery bronze above the trees topping the New Jersey ridges that march off to the West. The office, with its rich blond carpet, blond wood desk, and generous shelving, may be the most sumptuous room at the institutional, World War II-era Murray Hill site.
On top of Penzias's desk sit two computers, one of them with a camera on top; this is the Vistium system of which he is so proud. On the wall above it is a photo of the horn-reflector antenna that Penzias and Robert Wilson trained on outer space to detect the background radiation residue from the big bang.
Now Bell Labs's vice president for research, Penzias enters his office in time to chat with an assistant who is putting files, including several games, on an AT&T 486 notebook. Penzias will try them over the weekend; he confers with the assistant about one of them, a game called Shanghai that is based on mah-jongg.
This spirited interest in games is a happy sign. There were times during the last decade, when Bell Labs scientists felt they had to hide what amounted to technological game-playing from AT&T. There was Belle, the Bell Labs's champion chess-playing computer, retired in the early '80s. There was a ping-pong playing robot, a marvel of trajectory plotting, fuzzy logic, and space/time coordination that was quietly ditched because scientists feared stockholders and rate payers might grow cranky with such seeming frivolity. Bellcore's Bob Lucky, who moved into Arno Penzias's spot when the latter was promoted to vice president of research, recollected an even closer-to-home example of AT&T's unease with nonproductive science. When Lucky inherited Penzias's office, he found a plaque made from an AT&T advertisement produced when Penzias and Wilson were awarded their Nobel in 1978. "What does a Nobel Prize have to do with your telephone?" the ad asked. The questions troubled Lucky. "It didn't ring true," he commented. "Why should you have to explain why the discovery was important, why it was good for the telephone?" It was an important question, and a difficult one for Penzias, who, although clearly sympathetic to the need for intellectual elbow room, seems compelled to toe the AT&T market-driven-technology party line. In a discussion about Bell Labs's grand dilemma, Penzias is almost wistful as he reflects on the Unix Room and the highly consequential results of just hanging out. "One doesn't know when it's best to work," Penzias says, his precise English touched by an accent somewhere between Berlin and Brooklyn. (He was a youthful refugee from Hitler's Germany.) "It's hard to assign a time." I ask if it is still OK at Bell Labs to set your own time. "Reasonably so," he answers. "But not as OK as it used to be."
Not entirely satisfied with his response, Penzias tries out an analogy. "We've started operating more as ranchers than hunters and trappers," he says. "We used to run a trading post, where you'd come in and swap a good story for a grubstake." He goes on, "You'd get nitroglycerin, a long rope to lead your donkey, and you'd come back with nuggets or a good excuse."
Penzias, who has been at Bell Labs for 34 years, was an outspoken foe of the breakup of the Bell system. There is a residue of scorn in his voice when he talks about living "in a country where government declares victory when you buy from overseas." He maintains a certain longing for the era when it was not considered good form to let researchers "find out too much about development because they might get spoiled."
"Today," Penzias continues, "we have an environment of interaction;
while it is still collegial, you live with the business people." He
admits that at today's Bell Labs, "we want more process and
responsibility," but allows that "we have to be more careful - not all
searches are equal."
Not that this is a situation without advantages. Clearly, Bell Labs is still a scientific powerhouse. "One of our great luxuries," he admits, "is that we don't always have to be right." Another great advantage is that Bell Labs can invest in defensive research. "We don't want to be blindsided," he says, citing the example of a project investigating the neural nets of small animals to see if they are applicable to future chip design that more nearly approximates living intelligence. "Biology has a lot to teach us," he says. He also mentions some of the other marvels still pouring out of Bell Labs: the optical amplifiers and wavelength multiplexing technology that will broaden bandwidth to unimaginable degrees; the revolutionary digital-audio compression algorithms; the optical trapping technology that allows levitation and precise manipulation of matter down to the molecular level; the new type of semiconductor laser, the quantum cascade, able to emit light at fantastically specific wavelengths; and even the newly created lead-free brass alloy. Then there are the Nobel Prize winners of the future. But, finally, Penzias is a realist when it comes to assessing Bell Labs's role. "We're filling in a piece of the technology puzzle, an important piece but one not as grand," he says. Still, his belief in the importance of Bell Labs is as unassailable as it is unequivocal. "We're demonstrating that the investment in knowledge pays off in the long run," he says flatly. "America and the world would be a far different place without us." Innovation, Not Invention
An interview with Bell Labs's new president, Daniel Stanzione.
Punctuating the ongoing evolution at Bell Laboratories, 23-year AT&T veteran Daniel Stanzione is slated to take over as Bell Labs's eighth president on March 1, replacing outgoing John Mayo. A computer scientist by training, Stanzione's charge is to move the institution increasingly toward software development and away from the materials sciences that have historically been its strength. Wired asked Stanzione to comment on some of the important trends in science and at Bell Labs.
Wired: How has the focus changed at Bell Labs over the last decade?
Stanzione: It has gradually evolved as the industry has changed. Many of our competitors don't spend as much spotting new technology curves that will cause industry changes, but AT&T spends US$3 billion a year on R&D.
What are those curves?
There are three basic network areas: the broadband networks, the wireless networks, and client-server computing. If you go back five to ten years, you'd see that these three major network thrusts were not dominating attention, activity, or money as they are today. All of them are getting more funding. And in each, software is a key. Are there fundamentals that underpin all of those areas? They all involve developments in user interface. If you look at the history of these networks, you'll see that they were preceded by the invention of a user interface. Before there were telephone networks, somebody had to invent the telephone; before there were cellular networks, someone had to invent two-way radio handsets; before LANs, the PC had to come along; and before a new age in which broadband and wireless will allow networks to take off, someone will have to come up with new user interfaces. There are two other fundamentals: silicon and software. The AT&T ad that was done after Penzias and Wilson became Nobel laureates in 1978 asked the question "What does a Nobel Prize have to do with your telephone?" What is the answer to that question? Most technological advances that profoundly affect people's lives have, at their root, changes in the basic sciences. At Bell Labs, we have a dual obligation and also a dilemma: to make fundamental contributions for the betterment of all, and also to use technology for AT&T's competitive advantage in the marketplace. How do you do good and also do well? We need to do a better job getting that technology to market. You have to differentiate between invention and innovation. Invention is a wonderful quality of uniquely brilliant people but does not necessarily turn into real products and services. Innovation is getting technology to the marketplace. Never has there been an opportunity for technology to have such an impact around the world. If innovation is being able to get invention to the marketplace, then the golden age is still ahead of us.
/data/webs/external/dokuwiki/data/pages/archive/news/bcwired.txt · Last modified: 2014/08/18 00:26 by 127.0.0.1