(FORTUNE Magazine) – In Silicon Valley, venture capitalists often distill their collective business wisdom into maxims that belong on the bumper of a BMW or Range Rover. Bet on the jockey, they like to say, not on the horse. Find a parade and get in front of it. When the ducks are quacking, feed them. Cash is more important than your mother.

They don't always follow their own rules, however. Venture capitalists advise themselves, for example, to invest in product companies, not service businesses. They avoid capital-intensive industries dominated by corporate giants. And they look for business plans that promise leadership in small but rapidly growing new markets rather than in those that promise small shares of large existing markets. Which makes you wonder why VCs these days are financing so many startups eager to take on some of the nation's largest corporations in a business as old as the century, providing local telephone service.

In the past two years venture capitalists have put more than $80 million into fledgling competitive local-exchange carriers, or CLECs, which are vying with the regional Bells and independent communications behemoths known in industry parlance as incumbent local-exchange carriers (ILECs). Among the venture-backed newborns are Allegiance Telecom of Dallas; Covad Communications of Santa Clara, Calif.; Focal Communications of Chicago; MachOne Communications of San Jose; NorthPoint Communications of San Francisco; and Rhythms NetConnections of Englewood, Colo.

Behind the venture rule breaking are two changes--one legal, one technical. Because of the federal decree opening local markets, some of the $100 billion spent on local calls every year in the U.S. will inevitably go to new suppliers, and even a small piece of that market is a substantial business. In addition, the telecommunications infrastructure is in the midst of a historic transition from a voice network that can also carry data to a data network that can also carry voice. That transition provides an opportunity for new competitors in the services as well as in the equipment business.

CLECs aren't entirely new. The oldest--Teleport Communications Group (now part of AT&T Local Services) and MFS Communications (now part of MCI WorldCom)--have been around since the mid-1980s, when they called themselves competitive access providers, or CAPs. The CAPs connected directly to the long-distance network, enabling callers to bypass local phone companies and make cheaper calls. Because of the law at the time, however, the more successful CAPs built networks primarily in a few metropolitan areas where a small number of large corporate customers could generate enough revenue to justify multibillion-dollar network investments.

The Telecommunications Act of 1996 made a similar service possible for a new class of customers by enabling alternative providers to get up and running without having to build a network. CLECs now can lease parts of ILEC networks and even install their equipment in ILEC central offices. Thus Focal, for example, sells traditional local calling (and other services) to large corporations, and Allegiance does the same for medium-sized to small businesses. Both provide service through equipment of their own, which lowers costs.

The 2 1/2-year-old telecommunications act has also made possible new kinds of service for new classes of customers. Covad, MachOne, NorthPoint, and Rhythms don't even sell local phone service; they offer cheap, fast data connections between corporate computers and the Internet or remote workers. These embryonic CLECs are seeking to fill the gap between the bandwidth provided by cheap dial-up modem services and the expensive ISDN and T1 services offered by the ILECs. Although they call themselves CLECs, I prefer to think of them as phone companies without phones--data CLECs, if you will.

Covad, which last month filed for its initial public offering, was the first of this breed. Like the other data CLECs, Covad uses digital subscriber lines, or DSL, to deliver data at high speeds. DSL allows ordinary phone lines to carry information at 1.5 megabits a second, or about 50 times the speed of today's most common dial-up modems. That's equivalent to the speed of a T1 line that phone companies lease for between $1,500 and $2,500 a month. But Covad's DSL costs just $195 a month.

Rhythms is like Covad, but even more tightly focused on business customers, and it provides more extensive network services. NorthPoint supplies DSL access for Internet service providers. MachOne, which hasn't yet announced its plans, is aiming to provide inexpensive DSL access for homes.

I find the data CLECs appealing (I'm an investor in Allegiance, NorthPoint, and Rhythms) because it's relatively easy for them to get to the customers that count. By contrast, despite the promise of cable modems, cable television operators are unlikely to attract many business customers--cable goes almost exclusively to homes, not offices.

Ahead lies competition with older CLECs that already provide voice and data services, and with the Bell companies. But the latter, at least initially, are concentrating on DSL service for individuals, not businesses, in an effort to protect their lucrative T1 lines.

Most CLECs are losing money and will continue to do so for the foreseeable future. But that doesn't concern me too much. Cellular telephone carriers, cable television systems, paging networks, and alternative long-distance providers did the same in their infancies, before becoming huge businesses valued not on profitability but on cash flow per customer, to the delight of early investors.

That said, there are some short-term risks. Running a phone company isn't easy. Potential customers and investors should closely watch the most competitive markets, such as San Francisco, to see which contenders are doing the best job. Another risk is financial. Getting a CLEC up and running in enough markets to be viable takes about $300 million, and that's far more than venture capitalists ordinarily provide. To date, public investors have been eager to make up the difference through high-yield bonds and initial stock offerings. For example, although it had only one month of revenues from a single city and a mere $100 million in equity capital, Allegiance last January was able to borrow $250 million. But if the capital markets lose their enthusiasm for high-yield bonds, the CLECs have a problem.

For the near future, however--perhaps as long as five years--business demand for fast access to the Net will continue to outstrip supply. Demand will be far greater than the established service providers combined can meet, creating potentially lucrative company-building opportunities for almost any CLEC that can grab some land in this rush.

Tomorrow, the shakeout; today, as my venture capital friends say, a rising tide lifts all boats. Or, in a strong wind, even turkeys can fly.

RICHARD A. SHAFFER is founder of Technologic Partners, an information company focused on emerging technology. Except as noted, Shaffer has no financial interest in the companies mentioned. For an expanded version of Watch This Space, visit www.tpsite.com/tp/fortune/. If you have comments, please send them to shaffer@technologicp.com.