When Internet telephony first broke into public view in early 1995, it looked like a classic disruptive technology?an invention that would leave the mighty sprawled in the dust while exalting the small and fleet. The story was especially dramatic because some true Goliaths?the traditional telecom carriers?occupied ground zero.For all its trillion-dollar heft, the telecom sector seemed vulnerable because the price difference between the two types of telephony was obvious. A call?even an international one?cost nothing if it was made via a flat-rate ISP subscription that had been bought for other purposes, while a conventional call from the United States to Berlin or Hong Kong could easily run more than $100. Granted, the quality of IP telephony at the time was not high, but both cellular phones and speakerphones degraded quality too, and people had adjusted to those aural shortcomings.Journalists writing at the time saw no apparent end to the destruction. Long-distance companies would lose their toll charges. Governments would lose the surcharges set aside to support universal service. In addition, packetized voice integrates easily with digital services such as call distribution lists and interactive voice response. That means Internet telephony competed not just with conventional voice calls but also with the telco\u2019s ISDN offerings. All in all, the technology looked as if it was about to do major damage to the status quo.During the few years following, however, these stories turned out to be overwrought. Carriers dramatically lowered the costs of national and international long-distance calls. And the quality issues with Internet telephony turned out to be harder to tolerate than expected. While a real Internet telephony sector has taken root (according to TeleGeography, a Washington, D.C.-based research group, Internet telephony calls accounted for 1.6 percent of 1999 cross-border voice minutes), it is clearly not about to push AT&T, MCI and their ilk into the grave.But things continue to change. The quality problems of what came to be called voice over IP (VoIP) telephony disappear when voice packets travel through fast, well-managed networks. Accordingly, VoIP development has shifted from providing consumers with cheap calls over the public Internet to routing enterprise calls through LANs. There are several arguments for integrating voice and data. Communications comes under a single point of management. An all-IP environment makes it easier to scale old services and introduce new ones, such as user authentication or more sophisticated interactive voice response. Convergence is inevitable, so you might as well start the assimilation process now. During the past few years, a long list of VoIP-based products have mined that reasoning, resulting in software-based PBXs and Web-integrated calling centers to unified communication systems that combine voice, e-mail, fax and more. "The concept is to make voice just another network application," says Greg Ness, director of marketing communications at Shoreline Communications in Sunnyvale, Calif., a vendor of VoIP products.According to Ness, these products are especially appealing where telecom management costs are high, such as systems with a complex infrastructure (like that of a globally distributed services company) or that have to deal with highly volatile traffic loads, such as seasonal calls at travel agencies. For the moment, however, legacy systems still seem to rule the market. "At this point, I am not seeing any major increase in requests for VoIP projects or resources," says Raj Ganesan, a manager at KPMG Consulting in Radnor, Pa.If VoIP is not quite there yet, perhaps it\u2019s because it has not yet managed to live down the reputation it acquired in its early years, when voice packets traveled over the public Internet and suffered from the experience. Enterprises need to feel a high level of confidence before ripping up a vital function like telecommunications. What the technology needs is a big bet by a high-profile buyer or vendor, someone with a known commitment to call quality and a reputation to risk.Someone, in short, much like a telecom carrier. Beginning in early 2000, Global Knowledge, a major IT education company in Cary, N.C., noticed that its VoIP courses were beginning to fill up with carrier personnel. "We\u2019re seeing sales engineers, systems engineers, network operating people," reports David Mantica, a director at that company. He says carrier managers tell him that they see their future in terms of sales of IP-related services and are thinking that VoIP might represent the most economical delivery platform for those services. Oliver Hersent, CEO of NetCentrex, a VoIP platform vendor based in Paris, adds that new entrants see VoIP as a way of competing economically with incumbents and incumbents see the same technology as a cheap way of acquiring the capacity needed to handle a world of always-on phone connections.Thus, carriers may be saved by the very technology that was supposed to kill them, while packet-based telephony may end up being legitimized by the same companies it was supposed to bury. If any story could make futurists humble, this one should. Not that it actually will, of course, but it should.