Launched in 1989 under the leadership of Steven Sillens, then a manager of multiprocessors at Sun Microsystems, Livingston's networking drive began by creating a cheap router and communications server based on a new operating system, ComOS, specifically developed to help ISPs meet their Usenet burdens. Livingston quickly became a dominant force in Internet terminal servers and routers, and grew at a pace of more than 50% per year until engulfed by an explosion of demand in 1995. In August of this year, Livingston launched cheap low-end routers to serve both ends of an Internet connection: a $1,395 two-port PortMaster to link small offices to the Net at up to 230.4 kilobits per second and a sleek space-saving $3,495 PortMaster with 24 ports for ISPs. In October, Livingston announced a series of ISDN remote-access machines that will compete with the currently dominant Ascend ISDN pipeline system, if ISDN becomes the preferred mode of Internet access.

Now, everywhere in the Internet industry companies are resigning themselves to ISDN as the next "modem" (though, in fact it just brings into home and office the 64Kbps digital channels long used by the telcos between central offices). The scandal of U.S. telecom, however, is that the telcos could just as easily be bringing video capable T-1 service (1.544 megabits per second of bandwidth, equivalent to CD-ROMs) to homes if regulations permitted a reasonable tariff structure.

Moreover, new access technologies are emerging, such as cable modems and AT&T's new SDSL (symmetrical digital subscriber loop). Available this year and under test by Bell Atlantic, SDSL modems promise to bring T-1-Line capability to homes on twisted-pair copper wires for about $10 a month. SDSL follows many such copper prosthetics announced over the years (notably HDSL) [high bit rate digital subscriber line] from Level One, PairGain, Brooktree and others), all largely spurned by the telcos on pricing grounds, but capable of transforming the entire world of Internet access before ISDN's niggardly pipes catch on with the public.

While Internet hardware rushes ahead, Netscape, Sun and other providers of Internet software make the ISP a fast and elusively moving target for the telcos that wish to compete. With eight million browsers in the field, all upgradeable to the new 2.0 system--with the Java interpreter and Java multimedia programming language and toolkit--Netscape expects to attract some 100,000 software developers to its platform over the next year. There are already some 400 Java applications available, including word processors, spreadsheets and games that can play on any machine with a browser running a Java interpreter, regardless of operating system or microprocessor instruction set.

Netscape's expected army of 100,000 developers compares with some 10,000 developers for Apple's Macintosh and perhaps 3,000 for Microsoft's network, MSN. Emerging from a company that did not even exist two years ago, such a juggernaut will further empower the ISPs in their competition with the large invaders of the territory--not only the telcos but also the on-line services such as American Online and MSN.

The ISPs, however, are not usually in direct competition with the large phone companies. ISPs bring them new customers and new business users, and the ISPs also depend on them for home connections and for potential fiber-trunking services. The American telcos are currently laying some 1,300 miles of fiber-optic line every day. Moreover, beginning with TCI's and Kleiner Perkins' @home system, which functions with cable modems and new software from Netscape, the ISPs also may end up using cable plant. As cable modems become available, cable companies will likely turn to the ISPs to supply Internet services, local content, technical support and point-of-presence technology.

In the midst of these whitewater torrents of change, the some 4,000 ISPs and their increasing armies of supporters represent a serious threat to many of the established empires of telecom. Not only can they move much faster and more resourcefully, but they also have the key advantage of having bet exclusively on the PC and the Internet as the platforms of the future. However smart and powerful, Ray Smith, Mike Ovits of Disney, Gerland Levin and Ted Turner, Sumner Redstone and other aspiring Kings of the Road still entertain crippling visions of set-top boxes and interactive TV sets.

Andrew Grove of Intel had the last word for these efforts when he told Forbes ASAP last year: "By the time the set-top people reach the price points and form factors of consumer electronics and penetrate 30% of homes, the personal computer will be everywhere, controlling the TV like a minor peripheral." Bill Joy elaborated on this point in the October issue of Red Herring: "By the time [they] bring digital TV to the home, you will be able to take your Super Netscape version 4.0 Web browser with Super-Ultra-HotJava-Burners, and that will be your animated user interface. [The TV people tried, but[ it's like the Internet happened in the meantime. Right?"

Distracting most of the large companies (seen by the Internet's critics as impending monopolists), the pursuit of the set-top not only misses the point and begs the question but it also blows the key new hardware opportunity of the epoch. Although the PC will not be dislodged for most office applications, there is a real and rare chance today to create a new home architecture and software optimized for the bandwidth rather than for installed base. Together with the Java language, the Web browser breakthrough allows creation of new network PC and software architectures at price points that take advantage of the "hollowing out of the computer" caused by the impact of the Internet. Sun, Apple, Oracle and Jean-Louis GassŽe's BeBox are all focusing on this target today. All are trying to take advantage of the elusive opportunity of creating cheap machines optimized for bandwidth and graphics rather than for legacy software baggage (the storage can be supplied on the Net). That opportunity follows the PC and Internet model--the microcosm and the telecosm--into the cornucopian digital future of the information age, with the old analog TV and telephone left far behind.

Amid all these torrents of futuristic technology and prophecies of a tragic denouement in a wasted commons, it is comforting to return to the man who began it all, Vinton Cerf of MIC. Coinventor of the Internet protocol TCP/IP, developer of the once-pioneering and a philosopher of the Net, he is now in charge of MCI's data network, which includes MCI's Internet backbone network. A rare combination of technical grit and visionary enthusiasm, he faces resistance from forces within the company that still lust for the glamour of Hollywood and see the Internet as the CB radio of the 1990s. Nonetheless, Cerf at 52 is leading MCI toward a new Internet-centric strategy that is more likely than the MCI lobbyists to save the company from the grave perils of long-distance deregulation. The company is already creating a new backbone for the National Science Foundation part of the Internet, connecting supercomputer centers and other high-bandwidth applications at speeds of up to 622 megabits per second. MCI also is a major supplier of Internet bandwidth. Its network connects to all six NAPs (national access points) through which the ISPs link to one another.

Cerf observes that the national phone network grew at a similar pace through much of its history and regularly met ever challenge. The telcos, for instance, surmounted the predicted crisis of the NAPs early this year, when--following the withdrawal of government funds--the network was expected to collapse under galloping increase sin traffic. But the NAPs, despite unsuccessful struggles with the remaining instabilities of ATM (asynchronous transfer mode), ultimately rose to the challenge, saving the Net by using fiber optics and digitization, as well as transparent silicon and opaque silicon.

Today, new entrepreneurs are rising up to shape the future of broadband networks and possibly seize the market from the incumbent backbone suppliers. Silicon, both see-through ad solid, remains at the heart of the solution. One of the ways MCI is meeting the challenge of the future is by purchasing eight "gigarouters" from NetStar, a startup in Minneapolis that is exploiting Moore's Law to bring IP (Internet Protocol) switching and router technology into the microcosm.

Launched five years ago by a group of veterans of the Minneapolis supercomputer scene--Lee Data, Cray and other companies--NetStar went public this fall at a $83 million valuation. It is pioneering an elegant routing architecture that gets eight times the throughput of a Cisco 7500 at a 20% lower price. While existing routers run bits down shared backplane buses, NetStar's IP router reserves a full one gigabit per each of up to 16 media cards attached to a single-chip TriQuint 16 gigabit-cross-bar switch.

Ubiquitous on the Internet, Cisco remains an imperial force. But as the microcosm advances, it too faces threats. Not only can it not compete with NetStar at the top of the line but it also faces Livingston, Ascend and possibly even Compaq at the bottom.

Critics of the Internet have long predicted that as ever-more-turbulent floods of broadband data and Web images crowd the commons, the Net will no longer be able to bear the load. The routers in the NAPs and other critical paths will jam up and crash. But the microcosm enables a constant stream of exponentially more powerful new architectures as functions that were once spread out across entire boards collapse into single chips and multichip modules. For 1995 and beyond, MCI has bent on NetStar's feats of microchip integration to countervail every population explosion across the network commons. Following the laws of the telecosm rather than the megalithic visions of the critics, the fast new networks are becoming constantly dumber and more entrepreneurial. Ciena Corp., a small, venture-funded vendor of optical networks, is now supplying the next generation of back-bone gear, a system that can carry 16 separate bitstreams on every fiber thread. The first application of the new all-optical technology in public networks, is now Ciena's innovation is a precursor of the terabit (trillion-bit throughput) networks that will be filled with video teleconferencing, video on demand, virtual reality, and other bit-thronging and polygon-shuffling applications of the future.

Only one competitor, Northern Telecom, might challenge NetStar and the others providing the new superswitches dumb enough to prevail at the top of the line. In early October, Northern's BNR lab exhibited a terabit-switch architecture at the Telecom 95 show in Geneva. This machine, once again, illustrates the triumph of dumb networks. The dumb terminals of the past, whether POTS phones or mainframe 3270 panels, required smart networks, with central-office switches from Northern and AT&T containing no fewer than 26 million lines of software code. But the new Northern terabit uses passive optical components and virtually no software at all. It points to the evolution of a fibershpere for broadband wire traffic that will function like the atmosphere for wireless traffic. (See Forbes ASPA, "Into the Fibersphere," December 7, 1992).

While the critics of the new technology fix on the foibles of television and the monolithic aggregations of old media, the Internet is emerging as an entrepreneurial efflorescence. Comparing the Net to the decline of CB radio and the tragedy of the commons misses the providential convergence of the laws espoused by Moore and Metcalfe, with thousands of entrepreneurs in tow, exponentially expanding the commons with streams of new invention in a creative spiral of growth and opportunity. In seeing the technology as a killer of jobs and family life and a polarizer of opportunities between rich and poor, they miss the most radically egalitarian force in the history of the world economy.

The critics seem oblivious to the most basic realities of the U.S. job miracle. While the U.S. deployed three times as much as computer power per capita as any other industrial region, this country created some 45 million jobs in 25 years at rising, real incomes. Not only was the U.S. a world leader in the proportion of its working-age population with jobs, but it also created employment for some 12 million immigrants, while its corporations endowed new work for people around the globe.

At the same time, a billion people, mostly Third World Asians, used the technology to leap into Third Wave riches without ever having to endure a heavy industrial phase. Gaps between the rich and the poor collapsed everywhere that the networks reached, as former peasants around the world--from Bangalore to Los Angeles--gained new freedom and opportunity from the information economy.

The Internet creates jobs by making workers more productive, and thus more employable, regardless of where they live. By engendering more investable wealth, it endows new work, providing the key remedy for the job displacement entailed by all human progress. By aggregating distant markets, the Internet enables more specialization, and more productivity and excellence. It will help all people, but most particularly the poor, who always comprise the largest untapped market for enterprise. And the Internet will continue to grow, transforming the global economy with its power and building a new industry even larger than the PC's.

Fueling the transformation are the laws of the telecosm. They begin with Metcalfe's Law: The power of computers on a network rises with the square of the total power of computers attached to it. Every new computer, therefore, both use the Net as a resource and adds resources to the Net in a spiral of increasing value and choice. This means that any limited, exclusive or proprietary network will tend to lose business to a more open, accessible and widely connected network. Metcalfe's Law dooms all the dreams of the Time Warners of the world to create exclusive and proprietary combinations of content and conduit.

As a further rule, networks prevail to the extent that they feed on the invention and creativity of their users, since the power of the computers on the edge of the network will increasingly dwarf the intelligence of the network fabric itself. For example, a 5ESS central-office switch from AT&T, commanding some 10 MIPS (millions of instructions per second) and linking some 110,000 lines, once represented the most powerful computer in a local phone network. Today those 10 MIPS are infinitesimal compared to the collective computer power of the tens of thousands of personal computers, each commanding 20 to 100 MIPS, linked by modems to the switch.

Lacking an entrepreneurial environment of inventive users, the government-run PTTs (Post Telegraph and Telephone) of Europe have been rapidly losing ground to the U.S.'s more rivalrous RBOCs (regional Bell operating companies) and long-distance carriers, and all have been losing ground to the explosion of interconnected private nets. The U.S. has some 700,000 private networks compared to just 14,000 in Europe and some 75,000 in Japan. Private nets that feed on the creativity of their users will always tend to prevail over public nets, such as France's Minitel or American's interactive TV projects, that try to supply their entire system from a central office.

Eric Schmidt of Sun offers a true parable of the Net. Back when the Internet was the Arpanet, two routers were added to the system, but the routers' hopping ratio (the number of hops to any destination) got struck at zero. Because traffic always seeks out the optimal path, most of the traffic on the Net rushed to these two machines, since they promised instant transmission. Until the settings were corrected, the system was swamped.

On the Net, traffic will always gravitate to the most efficient broadband channels. If the telcos and software monopolists attempt to gouge customers in a badly designed and costly "top-down" network, traffic will migrate rapidly toward the freedom and bandwidth of a bottom-up solution. In the engineering global Internet, these channels could emerge among bypass suppliers using dark fiber; among low-earth-orbit satellite systems, such as Teledesic and GlobalStar; among cable companies and renegade long-distance suppliers; or among companies as yet unknown.

Guided by the valuations of the market, capital follows a similar rule: It is routed rapidly to the channels where it can be used more productively. At present, afflicted by perverse regulations that bar phone and cable companies from collaborating in the same region, valuations of these companies are low. Meanwhile, analysis complain of the excessive valuations for ISPs, such as Netcom, and their suppliers, such as Cisco, 3Com and Netscape. Not only traffic but also investment flows to the least regulated and most enterpreneurial arena.

A further law of the telecosm ordains that, in an age of dumb terminals and phones, traffic flows to smart networks full of intricate software. In an age of ever-multiplying computer power, impelled b Metcalfe's Law, traffic flows to the dumbest networks that gain their intelligence from the variety of powerful machines attached to them. A corollary is that, along with traffic, capital flows to the dumbest and most broadband nets with the most computer intelligence on their edges.

Perhaps most important of all is the cultural law of the telecosm. Networks promote choice, choice enhances quality and quality favors morality. Television is culturally erosive because its small range of offerings requires a broad, lowest-common-denominator appeal. Linking to millions of cultural sources, global networks provide a cornucopia of choices, like a Library of Congress at your fingertips. On the Net, as at a giant bookstore, you always get your first choice rather than a lowest-common-denominator choice. A culture of first choices creates a bias toward excellence and virtue.
The critics of the Internet are mostly skeptical about the value of choice. But choice validates freedom and substantiates individuality. Choice accords with the inexorable genetic diversity of humans. It makes possible individual aspiration and creativity. It is the lowest-common-denominator offerings of mass-broadcast media that lower humans to the animal level, eclipsing the differences that make us human, cutting off the higher aspirations and inspirations that elevate us beyond our appetites, reducing us to an impressionable crowd, zapping through the channels looking for a splash of blood or flash of nudity or demagogic spiel of hate.

In prophesying centralization and tyranny, the Cassandras miss the centrifugal forces of the Law of the Microcosm, overthrowing all monopolies, hierarchies, pyramids and power grids of established industrial society and endowing individuals with the power to be transcendent and free.    

TOP

Telecosm Index | Search | ASAP | ASAP Archives
Webmaster | Forbes | Gilder Tech Report

Copyright Forbes 1997