Ten years ago was the dawn of Voice over IP (VoIP). The pioneering Israeli company VocalTec had just released its VoIP software for PCs (it was named iPhone, BTW). Industry guru Jeff Pulver (whom I now partner with in FWD) had begun to hold his Voice on the Net (VON) shows. As the founder of VoIP startup ITXC, I was invited to give a keynote at VON in Boston.

The evolution of VoIP, I opined with the requisite PowerPoint slides, will be like a three stage rocket. I was right about the first two stages and dead wrong about the third.

The first stage will be fueled by arbitrage. VoIP provides a mechanism for avoiding the international toll booths of the telephony cartel. There is a huge amount of room to profit by providing cheaper service between countries since these services are now priced ludicrously above cost. We can turn what would have been international calls into domestic calls by routing the international portion over the Internet and resurfacing within the country being called.

This prediction was correct. VoIP was one of the major of the major contributors to the collapse of cartel pricing (called settlement rate agreements) for international calls. Over time (and with the help of deregulation), the wholesale cost of calls to places like India and China fell from dollars per minute to pennies. Easier communication played a major role in the economic rise of those two new economic giants.

The second stage, I predicted, will be fueled by technology. It no longer makes sense to have a separate network for the type of data called voice. Voice will be carried on the Internet like all other data. Routing will be much simpler when fixed connections between switches are replaced by the web of connections available on the Internet. Voice will both be more reliable (because the Internet as a whole is more reliable than a fixed network like the old telephone network) and much cheaper.

This, too, largely came to pass. Voice over the Internet WAS much cheaper than over leased phone lines even after the international settlement scheme collapsed. Even without an arbitrage advantage on many routes, the international portion of calls have moved increasingly to the Net. This trend was delayed some due to the overbuilding and collapse of communication networks of all kinds during the late bubble. Undersea phone cables weren’t as efficient as the Internet (actually often the same fiber but that’s complexity for another day) for carrying voice but they were priced down to near nothing after the bankruptcies of the companies that had financed and built them.

I concluded my prophecy by saying that the third stage would be fueled by innovative new phone services which VoIP would make possible. People would want a VoIP phone because it could do things that an ordinary phone couldn’t. The phone systems would become much better with VoIP at their core, not just cheaper but better.

I was wrong; the phone system hasn’t become better. The fixed handsets we use today and the services we get over them haven’t improved at all. VoIP services like Vonage (which I use) only make small improvements in things like voicemail and don’t change the way we make calls at all. It’s incredible, come to think of it, that we’re still dialing (or at least pushing) phone numbers and we have to know where a person is to reach him or her on a fixed line phone (not quite true with Vonage where you CAN move the phone).

Cellphones are somewhat better. At least they have directories on them and, when you call them, you are calling a person and not a place. You can chat on them, too, and send pictures.

With hindsight, I was wrong because the old phone paradigm can’t be incrementally improved in any significant way. Any major improvement is stymied by the need to keep compatible with the rest of the unimproved system. I also underestimated the ability of the entrenched phone companies to fight a legislative and regulatory battle against upstarts—they’re really good at that.

So what’ll happen? Are we stuck with the old calling model forever?

No way! There is no third stage of VoIP as an incremental improvement on POTS (Plain Old Telephone Service). Instead there is a whole new way to communicate. POTS won’t be improved; it’ll just be replaced. The rearguard action fought by traditional phone companies will eventually result in their over-priced and underperforming voice services being replaced and abandoned since they aren’t being improved (unless those companies can control the Internet—that’s a big unless).

In the new communication world—which is already forming inside social networks—live voice and voice mail are just two on a continuum of choices people have for communicating with each other. Video’s a choice; so is text and email and still pictures. Communication can be live and real-time; it can be slightly async like texting; or seriously async like email. The modes of communication mix freely. Two or more people using different devices communicate at the highest common denominator rather than the lowest.

And there are no more phone numbers, just names and handles (made up names). There’s no more great directory in the sky; there’s the union of the directories of the social networks we use and our personal directories. We’ll know who’s “calling” us as surely as we know whom we’re calling (callerID today tells you where a call is made from, not who is making it).

Location will be largely irrelevant (as it is with email) except in special circumstances when we want to make it relevant—“everybody in the flood plan of such and such a river should seek higher ground ; “the emergency call is coming from this location”. Location is cyberspace will be equally as relevant or irrelevant as geographic location. “I’m in the such-and-such network on Facebook or I’m a fellow online Rotarian” are relevant uses of cybergeography. But friends’ lists won’t stay restricted to single social networks like Facebook or MySpace for long.

Oh yeah, the rocket that’s come in from left field to replace the dud on top of the VoIP-as-POTS-replacement three-stager doesn’t know anything about charging by the mile or by the minute or even by the “call”. It’s sort of like email that way.

It’s been an interesting ten years. And it’s only the beginning.