THE RACE TO REWIRE AMERICA Communications giants want to build networks that will change how you live. They're making billion-dollar bets. Their future -- and yours -- is on the line.
By Andrew Kupfer REPORTER ASSOCIATES Alicia Hills Moore and Antony J. Michels

(FORTUNE Magazine) – Call it the first great business showdown of the 21st century: The giants of American communications are locked in a struggle to build and control a vast web of electronic networks. These so-called information highways will be of glass fiber and will deliver an abundance of services to offices and houses -- video images, phone calls, helpful data in many guises. They promise to change the way people work and play. In the view of some technologists, they could affect American life as profoundly as railroads, interstate highways, telephones, and TV. The risks are as colossal as the opportunities. Building a glass highway is moon-shot expensive -- by one estimate, extending the networks over the next 20 years may cost phone companies alone nearly $140 billion. Regulations are likely to change while the game is being played; technology is evolving so quickly that some highways could become obsolete before they are complete. The highways' success will depend on the revenues they generate. Yet no one knows how much consumers will pay to browse through movie libraries using their remote controls, play electronic games with far-off friends, or visit their doctors by video. Since they control the information conduits feeding into households today, telephone and cable TV companies have the most at stake. The idea of receiving a phone call on a cable TV wire may sound as impossibly counterintuitive as, say, getting a cup of coffee out of an electrical socket. But the idea will soon be reality. New technology is breaking down the barriers between the industries. Telephone companies such as Bell Atlantic and GTE are eyeing the lucrative $20-billion-a-year cable TV business, while cable operators such as Tele-Communications Inc. and Time Warner (parent of FORTUNE's publisher) covet ; the vast $65-billion-a-year market for residential phone service. The highways these rivals have started constructing are different from the electronic superhighway that Vice President Gore is promoting. His is a national network of supercomputers, linked by fiber optics, that will connect universities, hospitals, research centers, and other institutions that need to exchange vast amounts of data. Construction of this superhighway, an expansion of today's federally subsidized scientific networks, seems almost certain to proceed. But the Clinton Administration is counting on private enterprise to construct advanced networks that will serve the public generally. As a result, America's information system won't have a single owner: It will be a network of networks, controlled by many companies. How government regulates the networks -- or doesn't -- will profoundly influence service and profits. Should regulators scrap rules that prevent cable TV and telephone companies from leaping into one another's businesses without constraint? Should government take the lead in ensuring that the networks all work together? If not government, who? The most controversial question is whether business, without the help of Washington, will act quickly enough. Many people fear that the U.S. is lagging dangerously behind its trading partners in building information highways -- a failing that could reduce America's competitiveness. Corning, the No. 1 maker of optical fiber, estimates that if telephone companies upgrade aging installations at their historical pace, the rewiring will take until 2037. But Japan is committed to completing a national fiber network by 2015 and believes that the resulting productivity gains will boost GNP by no less than 30%. Germany and France are not far behind in their plans. Observes Michael Morrison, manager of advanced operations testing at GTE: ''These nations see how attracting and keeping companies with telecommunications helps them be competitive. We tend to trip over our own feet.'' Amid the debate and uncertainty, companies like Morrison's are placing billion-dollar bets. What follows is a look at the policies and technologies that will shape the new highways, the services they will make possible, and the competitive strategies of those who plan to build them.

NEW RULES OF THE ROAD -- The Clinton Administration has put electronic highways on the national agenda but has yet to decide what Washington will do to get them built. One of the hottest debates is about how much the government should spend. Some people think zero. Brendan Clouston, chief operating officer of Tele-Communications Inc. (TCI), asks, ''If multiple industries will create the architecture, why should taxpayers pay?'' But the highway builders face a chicken-and-egg problem that a sprinkling of government seed money could help solve. Unless fiber-optic networks can provide services that consumers want to buy, they will be just so many useless strands of glass. Useless, expensive strands. Meanwhile the businesses that might offer services such as movies on demand will need to invest in specialized hardware and software. Unless networks stand ready to carry their services, making these investments would be like putting up a motel before the road is built. Lee Camp, president of Pacific Bell Information Services, describes the dilemma his company faces: ''Does one pursue a Field of Dreams strategy -- 'Build it and they will come' -- or wait until there is proven demand?'' A modest investment of federal dollars could kick-start the industry, argues economist Eli Noam, a Columbia University professor and authority on telecommunications. Demonstration projects electronically linking, say, community libraries, schools, and universities could pique public interest and stimulate demand for high-capacity networks. Consultant Janice Obuchowski, a former Bush Administration official, advocates ''applications funding,'' including grants to help entrepreneurs develop services to sell. The Defense Department's Advanced Research Projects Agency or a civilian equivalent could also contribute to key technologies. In the 1980s the agency helped finance work on digital signal processing -- the packing and unpacking of information for efficient transmission and reception. The technology still needs work. So does the science of translating video images into computerized form. Regulators, meanwhile, could hurry the highway by loosening antiquated rules, especially those that hobble the telephone business. For instance, phone companies must depreciate their capital equipment over 20 years or more. That was the useful lifetime of telephone gear a decade ago, but today, when technology changes faster than governments in Bolivia, the rules deter investment. The Federal Communications Commission and state utility commissions, which must arbitrate the networks' construction, are besieged by lobbyists from telephone and cable companies. Each side wants to gain an advantage. While the regulators deliberate, cable companies are getting into the phone companies' business and vice versa. Some cable operators have bought into companies that are laying fiber in city centers and stealing customers from the telephone network. A Bell company, in turn, just made an end run around rules barring it from the cable business in its service area by acquiring two big cable franchises out of state. As cable and phone companies invade one another's markets, the government should dismantle the rules separating the industries carefully. So far the phone companies have won the right from the FCC to offer a ''video dial tone'' that enables other companies to use the phone network to transmit video programming. A few phone companies have also won relief from so-called rate- of-return regulations, which they say stifle innovation. The rules, administered by the states, limit the companies' return on assets, typically to 8% to 11%. If a company takes a chance on an ambitious new service and succeeds, it must return any ''excess'' profits to customers in the form of rate rebates. Some states, such as New Jersey, have wisely begun to allow higher profits in exchange for guarantees that a new investment will lead to lower prices for basic phone service. The most difficult issue government will face is how -- and even whether -- to make sure there is basic, low-cost service for every American who wants a phone and other essential services that the highway will provide. On the telephone network, that principle, known as universal service, has been the law of the land for 60 years. It reflects the belief that phones, like mail, electricity, and highways, unite the nation's people and make America strong. The government achieves universal service through regulation. The phone companies are obliged to hook up everyone in their service area and charge each customer the same basic rate -- even though this can mean stretching miles of wire to a customer whose payments won't cover the whole cost. High profits from some customers -- such as those who pay for added services like call waiting -- compensate for the money losers and enable the phone company to hit the rate of return the regulators allow. But as competition and new technology galvanize local markets, universal service becomes harder to deliver in the traditional way. Cable companies aren't bound by universal service rules. Using leading-edge technology, a savvy cable operator could add phones to its system and target just those parts of the local market that produce the fattest profits. Indeed, Time Warner has asked the FCC to let subscribers on its advanced cable system in Orlando, Florida, use the cables to place long-distance calls. That would enable the company to claim a share of the lucrative fees that AT&T and other long-distance companies pay for connections to local customers. If the FCC approves Time Warner's plan, BellSouth would see its profits in Orlando erode -- and might file for permission to raise basic rates. What should the FCC do? Option A is to regulate local markets more, requiring newcomers to provide universal phone service. That would surely discourage competition and slow the development of information highways. Option B is to lift restrictions on phone companies just enough to let them counter, but not drive out, the invaders -- say, by providing information services, including their own video programming, as long as such services are fairly priced. This approach involves a delicate balancing act that federal and state regulators would have to perform again and again as competing local networks pop up across the country during the coming decades. Still another option: Require new entrants to pay the phone company to help offset the cost of universal service. Opening local markets to competition will be a difficult business because of the complex and interlocking nature of regulations affecting the industries -- far trickier than opening up long distance, a process that has occupied the FCC and the courts for more than ten years. Eventually, in communities where competition flourishes, regulation may not be needed to ensure cheap and abundant telecommunications. In the meantime the government will have to be careful to modify its rules in a way that protects the public interest while giving neither industry an unfair advantage. Letting many companies compete in building the information highway lessens the chance that the country will get married to the wrong technology. Competition will foster continuous innovation. But it also increases the risk that the U.S. will be dotted with networks that can't talk to one another. Ah, for the simplicity of monopoly. When AT&T ran the Bell system, it kept everything working smoothly by setting detailed technical standards. When a new service appeared, it was sure to work everywhere. ''But boy,'' says Bob Barada, chief strategist for Pacific Telesis, ''was that process slow! This country can't wait for a standards body to cross every t before we get started.'' Instead of defining standards in advance, regulators should jawbone companies into working out the details themselves. If a cable company and a telephone company operate competing networks in a community, residents should be able to reach each other no matter which network they use. But where should the physical connection between the networks occur? In a manhole? In a telephone company central office? And with what equipment? To be maintained by whom? These are nuts-and-bolts questions that regulators aren't good at answering. Columbia University's Noam, who once served on the New York State Public Service Commission, advises regulators to bring all the parties into a room and tell them to work out their differences under threat of regulatory fiat. He says, ''The arrangement works particularly well when it involves technologists -- they're problem solvers.'' Once the rules are agreed on, regulators can codify and enforce them. If they do this job well and clear the way for competition, says Jim Chiddix, the chief technologist at Time Warner Cable, ''it looks like all the forces are there for promoting tremendous innovation in technology: fear and greed.''

THE GLASS HIGHWAY -- ''You'll be going to Cerritos, I hope,'' said the man from GTE. Cerritos is GTE's community of the future. Buried beneath the wide, straight roadways of the Los Angeles suburb, slender sheaths of glass guide pulses of infrared light from lasers in the switching center to two schools and 4,200 homes, bearing programs and telephone calls. A teacher summons up video lessons at the touch of a button. Some families on the network -- brave Jetsons! -- can call up movies on the system whenever they wish. The families can even converse with each other on the screen. All two of them. ''For years it was only one guy watching movies. It's a standing joke in the industry,'' says Danny Briere, president of TeleChoice, a New Jersey consulting firm, of the cautious pace at which GTE has pursued its five-year trial. But Cerritos is no joke to GTE. Billion-dollar decisions depend on what technologists and marketers learn there; other companies are conducting similar small-scale trials. They show the fitful and tentative way revolutions start. The technology of the information highway is evolving at a furious pace. In December 1992 the FCC licensed CellularVision, a Freehold, New Jersey, startup company, to test an ultrahigh-frequency microwave radio system that may eliminate wires in some parts of urban networks. Such innovations could dramatically lower costs and reshape information networks even as they are being built. The highways that cable and telephone companies currently envision will, in the words of GTE vice chairman John Segall, ''tie the world together in a hush of photons.'' The network will be rich in fiber-optic cable, which has far greater carrying capacity than copper wire or coaxial cable. Messages conveyed on the fibers will be encoded in the ones and zeros of computer language and compressed by sophisticated circuitry for easier storage and quicker transmission. Ultrafast switches will route video images as easily as ordinary phone calls. Special computers called video servers will store movies and TV programs in digital form. These technologies will give the network its hallmark attributes. It will be ''broadband.'' Just as a line painted with a broad brush contains more paint than a line traced with a narrow one, a broadband network can carry more information than its narrowband counterpart. Since signals on the network will all be digital, it will easily carry information of different kinds: It won't need to know whether a transmission describes a lark's song or a slasher movie. The network will also be two-way and interactive: Every user will be able to send all kinds of information -- voice, video, data, and graphics -- to anyone else. Before this vision can become reality, phone and cable companies must each overcome innate weaknesses. Phone companies are experts at running networks linked by switches (powerful computers that let any customer dial any other) and at providing service with near-total reliability. But the system itself is narrowband, its thin-gauge copper wire unable to carry a high-quality video image. Cable systems, with their heavier-gauge coaxial cable, are broadband -- a strength. But unlike telephone communication, which is two-way, cable signals flow in only one direction on the systems common today. They have no switches and can't relay phone calls. GTE's experiment in Cerritos typifies the approach phone companies will probably adopt. A fiber strand runs from the phone company's central office to a curbside pedestal that can serve up to 20 houses. Inside the pedestal is an / optical interface unit with a separate circuit card for each house. The card contains the subscriber's coded address and ensures that phone calls and video programs arriving on the shared fiber-optic line end up in the right place. The circuits also convert the incoming light pulses into electronic signals, which enter the household via coaxial and copper wires hooked to the TV and phone, respectively. Cable companies, by contrast, don't need to take fiber all the way to the curb. They will run it to the edge of each neighborhood, where transmissions will feed into the coax network that is already in place. Each fiber link might serve as many as 2,000 families. By using the latest compression techniques, which can multiply tenfold the number of channels on a cable system, a company can assign channels to individual customers as needed -- to deliver a movie, say, or relay a telephone call. The fiber links are essential for two-way communication; coaxial networks alone can't handle it. In a coax system, signals pass through an amplifier every 2,000 feet or so. Each introduces a whisper of electronic interference to the line. In one-way transmission, the noise is manageable; but on the return path in two-way communication, it builds up, and the cacophony of the amplifiers drowns out the message. The introduction of fiber brings a measure of calm: Laser signals can travel for miles without a boost, so the total number of amplifiers in the system stays relatively low.

For both industries, the most expensive job will be laying down fiber. The work has barely begun. According to Corning, the U.S. now has some 12 million miles of fiber installed -- compared with 1.2 billion of copper phone wire. Neither phone nor cable companies have put down much fiber in residential areas, which account for some 65% of the mileage of telephone networks and 75% of the mileage of cable systems. It's hard to say how much rewiring for advanced networks will cost, partly because both industries are gradually switching to fiber anyway for their ordinary operations. Corning estimates that doubling the rate of conversion of the phone system -- which would mean that the job would be finished by 2015 -- would increase spending over the period by $24 billion. Add the $63 billion that phone companies already plan to spend and $50 billion for new ultrafast switches to keep traffic flowing smoothly on the information highway, and the bill comes to $137 billion. That's just for the telephone network. Until now, cable companies have held a theoretical advantage: They can make do with less fiber because they already own a broadband conduit into the home. By most estimates, a cable operator could add two-way services, including fiber to the neighborhood, for less than $1,000 per household. Installing a Cerritos-like system, including fiber to the curb, could cost its telephone rival hundreds of dollars more. That advantage will erode: As demand for two- way services increases, the cable operator will have to segment its network into smaller units and install more fiber. Eventually the two systems will look and cost just about the same. But in a developing market, the cable company's head start might be crucial. Small wonder that Bellcore, the research arm of the seven regional Bell operating companies, has raced to find a practical way to transmit TV programs over ordinary copper telephone wire. In June 1991 it unveiled a digital compression system built to do just that. Known as ADSL (asynchronous digital subscriber line), the technology is still far from perfect. The longer the copper pathway the TV signal traverses, the more the picture degrades; at best the picture quality is no better than that from an ordinary VCR. All the same, if regulations allow, phone companies can now look forward to offering video service as soon as they bring fiber to within a mile or so of a residential area. Both industries need more breakthroughs. The greatest technical roadblock involves storage technology. In most electronic-highway plans, TV watchers will be able to scroll through menus of video libraries -- Treasures of Columbia Pictures, say -- stocked by independent vendors. A push of a button on the remote control and the show will begin. For such schemes to work, any company that wants to offer a video service should be able to buy a video server and hook it into the network. But servers that can store movies digitally and dish them out on demand aren't ready yet. The task is crushing: Even a 95-minute film like Wayne's World requires billions of bits of memory. Ameritech is testing a system in Chicago that will enable Arthur Andersen, the consulting firm, to dispatch training films to clients. Other developments, such as the rapid evolution of wireless technology, could change the course of the highway race. In CellularVision's trial installation in Brooklyn, New York, subscribers with a decoder box receive 50 - TV channels using a movable antenna only five inches square. Unlike ordinary microwave signals, which require a direct line of sight between transmitter and receiver, the ultrahigh-frequency signals bounce off concrete like a billiard ball off a felt rail, losing very little strength. So users need only move their antennas around until they get a good bounce. The system can also carry signals both ways; a test of telephone service will begin shortly. Most telephone and cable executives dismiss the idea that any new technology, no matter how startling, will provide its owner with a sustainable edge. In the long run, the same technology will be available to all comers. TCI's Brendan Clouston gets downright testy when pressed on the pros and cons of competing schemes for an advanced network. ''I don't like the way this conversation is going,'' he says. ''Technology is not the issue. What do consumers want to buy? What do they want to pay, and when?''

THE KILLER APP -- As they imagine the billions of dollars consumers might spend on electronic highways, telecommunications executives often exhibit the Pavlovian response of a gambling addict exposed to flashing neon lights. Listen to Arthur Bushkin, president of Bell Atlantic Information Services: ''The marketplace is not the $20-billion-a-year cable market or the $12-billion-a-year movie rental market. It is into the hundreds of billions of dollars. It's for work force training, medical services, and shopping. It's the ability to see real estate before traveling there. It's videoconferencing and using multimedia. It's transmitting recipes. It's endless.'' Maybe it will be, someday. But creating services that consumers will want to buy could make building networks seem as easy as running a string between two tin cans. Some applications are no-brainers -- merely better ways of delivering services people already use. Others, the kind visionaries cite when they claim information highways will change the American way of life, pose obstacles to execution and problems in predicting how consumers will respond. Almost without question, business demand will drive the market for advanced services at first. Manufacturers and their suppliers will use electronic highways to link their computers and collaborate on product development. Insurance companies could receive images of auto wrecks for claims processing. Video depositions and arraignments, which some law enforcement agencies already employ, would become common. But to really cash in, communications companies will have to turn consumers on, says John Malone, president of Eastern Management Group, a Parsippany, New Jersey, consulting firm. Just as the personal computer industry languished until spreadsheet programs appeared, information highways need a ''killer app'' -- software industry lingo for an application people are dying to use. The No. 1 candidate for killer app status is video on demand, the armchair equivalent of a trip to the perfect video store. Viewers will be able to order movies and TV shows anytime, using remote control. TCI recently studied how such a service would compare with today's more cumbersome pay-per-view, which requires customers to phone in their orders. It found that viewers would increase movie spending three to five times. Another likely hit: telecommuting. Despite the limitations of today's telephone networks, the number of employees working at home is rising at a startling rate. According to Malone, 14% of the FORTUNE 500 and Service 500 companies now have formal telecommuting policies; 870,000 employees work at home 35 hours or more each week, and 5.5 million do at least some home work. The numbers are growing by more than 35% a year. The advent of information highways will accelerate the trend by increasing the number of jobs telecommuters can perform. An American Express service representative, for example, wouldn't have to leave home to field customer calls and tap into the company's immense databases. Telecommuting will get a lift in states like California, which requires companies to encourage the practice as a way of reducing auto pollution. The educational possibilities of the advanced network are emerging in tests. Betty Hyatt, a teacher in Cerritos, uses the fiber-optic system to call up penmanship lessons for her third-grade class. That frees her from the chalkboard so she can roam the room and monitor her pupils' progress. Hyatt says, ''It's changed the way I teach.'' Ameritech has begun a program in Warren, Michigan, that will link the homes of 115 fourth-graders to their classrooms, allowing the children to call up their homework electronically and do it on-screen. Advanced networks will eventually let students in remote areas attend college classes by wire. And they may matriculate not at Ohio State but at the Big Ten, mixing and matching video classes from any of the member universities. For physicians, the house call may return, electronically. Using a video link on the network, a patient could see and talk to her doctor without leaving home; by placing a hand on an electronic sensor, she might relay vital readings the doctor could analyze. The highway may be dangerous for debtaholics. Going on a buying spree will be as fun and easy as playing videogames, with no need to sit like a brass monkey before the Home Shopping Network. An armchair consumer will select a video catalogue from the on-screen menu and, by punching the remote control, ask to see a jacket in a certain size and color. A simulated three-dimensional model will rotate slowly on the screen. The subscriber can order by pushing a button; the network will have his address and credit card data on file. Marc Porat, head of General Magic, a Silicon Valley software developer, believes the advanced network will change the way people buy information. He expects a form of publishing to emerge called electronic subscriptions. It will replace the sort of books that become obsolete as soon as you buy them -- guides to New York City nightlife, for example. A broadband network could deliver an update every month by either displaying it on screen or transmitting data to the consumer's home printer. Eventually the highway may deliver a lot more. A jovial, forward-thinking engineering colleague of Porat's predicts that people will don electronic gear and use the network to play virtual reality games. Players will have the illusion of moving through an artificial but lifelike 3-D landscape. That may put a new spin on humanity's oldest killer app. ''The ultimate,'' says the engineer, ''is when you'll be able to put on a visor and bodysuit that let you become anyone in the world having sex with anyone else in the world.'' Virtual reality enthusiasts call it teledildonics. Virtual reality is the most extreme variant of so-called multimedia programming. In partnership with GTE, the Discovery Channel is already transmitting coded instructions in its TV signal that add graphics to the station's science and nature programs when they appear on the Cerritos system. Viewers may see a map superimposed on the screen, or a fact about an animal habitat. The information comes from a miniencyclopedia on a CD-ROM that plays in a device connected to the TV; the codes in the TV signal summon up relevant bits during the show. Only 50 or so families receive the disks now; nationwide rollout may begin next year. When advanced networks are ready, the disks will be unnecessary; the extra information can travel over fiber. But like other services upon which the information highway will depend, multimedia is having trouble getting off the ground. Production is awkward and enormously expensive; most CD-ROM programs are as primitive as the Groucho Marx Show in TV's early days. That's one reason network builders are having a tough time predicting customer interest. TCI's Clouston says: ''Traditional marketing is done with products that exist. What we're doing now is like asking a horse-and-buggy driver whether he'd pay $100 more for a car with an air bag. He'll ask, 'What's a car?' Nobody knows what people will spend.'' That leaves cable and phone companies with high hopes and gnawing doubts, like city leaders who erect a lavish sports dome in hopes of landing an expansion team. Are the network builders in for a nasty case of that queasy feeling you get when you wake up in a bad, bad place?

GIGADOLLAR GAMBLES -- Gaining an advantage in 21st-century telecommunications won't be cheap. Stewart Personick, a networking information services executive at Bellcore, explains the cost of simply entering the race: ''If you want to make a commitment, you have to have a million customers. The investment in optical fiber, network hardware and software, automated billing, and advertising is a minimum of $1,000 per customer. You've got to go for a billion dollars.'' If technology, regulations, or business relationships change unexpectedly, that billion could vanish. Yet companies that hesitate could lose out completely. Even with imperfect technology, a big enough player making a big enough bet could stake out a dominant position. As a result, says Personick, the competition is like an Olympic bicycle sprint: ''All of these guys are on their expensive racing bikes going five miles per hour, waiting for someone to make a break. And then they all go like mad.'' The break has clearly begun. Last October, TCI announced it would soon begin offering some subscribers 500 channels. In January, Time Warner upped the ante with its Orlando plan to build a two-way network for 4,000 families. Then Cablevision promised a similar system in the New York metropolitan area -- for over a million subscribers. Meanwhile, US West unveiled plans for fiber-to- the-curb systems in its 14-state region; Bell Atlantic won permission to replace with fiber all the copper wire in New Jersey by 2010. Virtually every | other big communications company has an information highway plan in the works. Among the contenders, Bell Atlantic stands out for aggressiveness and astute politicking. Its plan to rewire an entire state is a first. It convinced regulators that New Jersey needs the expensive new systems to maintain competitiveness. Bell Atlantic wants to time the installation of fiber to suit local markets in each of the seven states it serves. In a few neighborhoods, where marketers believe there is ready demand for interactive service, the company is extending fiber to the curb right away. That has thrown a scare into at least one cable company. The owner of a housing development in northern Virginia says, ''When the cable people found out Bell Atlantic was putting fiber optic in, they had a fit.'' Where the company thinks TV watchers will embrace an alternative to cable, it plans to take fiber to the neighborhood and send ADSL transmissions the rest of the way over existing copper wires. US West is betting that the fastest way to roll out advanced networks is by cutting costs. The company has challenged suppliers to tighten their belts and help it build fiber-to-the-curb systems in new neighborhoods for no more money than a standard copper-wire system. (Right now putting in fiber costs about 30% more.) TCI, the largest cable company, will switch to digital signal transmission in dozens of communities starting next year and has ordered one million state- of-the-art converter boxes to let customers tune in. The investment could serve as the foundation for two-way networks that can deliver video on demand. Archrival Time Warner has focused its attack more narrowly, concentrating on showcase projects designed to push network technology as far as it will go. The system planned for Orlando will have 600 digital channels for video on demand and phone calls, as well as 75 regular channels for ordinary TV. Construction is set for next year, even though crucial components such as video servers aren't yet available.

GTE, which owns local systems in 40 states, is maneuvering to cash in on a key advantage over other regional phone companies. Since it was never part of the Bell system, GTE is not bound by the federal consent decree that bars Bell companies from owning information businesses. GTE is testing interactive video services it can market through systems like the one in Cerritos. One lets customers pay bills by filling in on-screen checks; another helps students prep for SAT exams. AT&T, finally, is poised at the edge of the field. It no longer owns a wire into the home, but with its proposed $3.8 billion investment in McCaw Cellular Communications, it could again be a powerful force in local markets, especially if wireless technology emerges as a way of delivering advanced information services. As they circle one another warily, phone and cable companies are like predators at a jungle water hole, wondering, Will it try to eat me, or will it kill some other animal and let me share the meal? Conflicting motives leave them torn between competition and cooperation. Phone companies could easily afford to put broadband wire into people's homes -- regulators permitting. But the job would take a long time, and the companies lack experience in programming. Cable operators are in a position to skim off lucrative telephone business. But they have little experience with network management, and none with switching or billing phone calls. The industry's fragmentation also suggests a need to cooperate: Most metropolitan areas have several cable systems, which would probably have to work with the phone companies to provide local service. Temporary alliances are taking shape. In Denver, for example, AT&T, US West, and TCI have teamed up to test-market video on demand. Viewers at home browse through a catalogue of 2,000 movies and punch in a code number on their remote controls. Exactly five minutes later the film starts playing, as if fetched from a computer's memory bank -- though what actually happens is that a worker at the test center jumps up, finds the proper videotape, and plays it in a bank of VCRs (below). Customers pay $4.99 per showing, about $2 more than for a rental cassette. Which companies clash will depend partly on the speed of deregulation. If state regulators move slowly, phone companies may go outside their service areas to get into new businesses, invading one another's turf. Southwestern Bell, for example, recently bought two cable systems near Washington, D.C., putting itself on a collision course with Bell Atlantic. Where regulators are most flexible, phone companies will simply upgrade their own networks. They may even ally with local cable operators to economize. In such an arrangement the phone partner would handle switching and billing for calls and interactive services delivered through the cable system. Eli Noam of Columbia University thinks regulators should be leery of such , plans, lest they give rise to powerful monopolies in local service: ''Phone and cable companies should be beating each other over the head in their home service areas.'' But Clouston of TCI argues that alliances would speed network building. In fact, technology is evolving so quickly that monopolies seem unlikely. CellularVision, with its capable little antennas, has sent strategists scrambling at phone and cable companies alike. The household antenna and decoder box cost only $300 to install, much less than any glass highway hookup. Such innovations could alter the balance of power. Alarmed, several Bell companies tried to squelch CellularVision's license application by filing objections with the FCC, saying the technology wouldn't work. But it does. During a recent demonstration in Brooklyn, the picture quality was good, except when a moving crane passed before the window. Among the would-be builders of information highways, it's too early to pick winners. At a recent press conference about Time Warner's ambition to build glass highways and fill them with photons bearing movies and recipes and homework, CEO Gerald Levin showed a flash of grim realism. Reflecting on the company's multimillion-dollar losses on Time Teletext, an early electronic information service, he said: ''A lot of people are going to lose a lot of money.'' No doubt. But the bounty will be great for those who marry the right technology to the right services at the right time.

CHART: NOT AVAILABLE CREDIT: NO CREDIT CAPTION: STILL A LONG WAY TO GO Companies with the spending power to build information highways include big names in telephones and cable TV. All employ optical fiber but have installed almost none in neighborhoods where the networks are densest. Phone systems still rely on copper wire, cable systems on coaxial cable.