Relying solely on the models of the past as predictors of the future can leave even the bestreasoned business plans all wet.

If we had to do it all over again, says the inventor Dean Kamen, we wouldn’t distribute water the way we always have. We wouldn’t rely so much on centralized collection and processing points, but instead would move water collection and treatment closer to the end user.

2010-StewartSchleyKamen’s thesis is that safe drinking water can be created on the spot with the help of a machine (one he happened to invent) that effectively eliminates, or at least reduces, the need to rely on massive distribution networks of aqueducts and pipes.

Kamen’s purification machine, the Slingshot, is an affront to centuries of water distribution practice tracing back to ancient Persia, where sloping tunnels called “quanats” were carved into hillsides to coax water down into channels built around cities, and later to Rome, where among the Roman Republic’s most impressive feats was the construction of the Aqua Marcia aqueduct, a slithering channel that laced its way across and under hills and earth for 23 miles before delivering water to public baths and fountains.

Similar approaches to moving large amounts of water would come to prevail in colonial America and beyond. Boston is credited with incorporating the first public waterworks system, in 1652. The Croton Aqueduct, which carried water from Westchester County to New York City, was completed in 1842. Chicago devised an underground tunnel system for distributing water in 1869.

With these delivery networks established, the next big leap was purification. In the early 1900s, American and European scientists and engineers began to devise largescale purification systems that protected waterways from pollutants and purged deadly waterborne contaminants from the public water supply, making it safe to drink distributed water without fear of contracting dysentery or typhoid fever.

The endurance of centralized, networked water delivery testifies to its utility and practicality. But occasionally, technology and invention rise up to challenge common thinking. That may be happening in a micro sense with Kamen’s Slingshot, which Kamen insists can create potable water out of just about any liquid source – no matter how tainted.

There are interesting parallels here with modern communications networks, which seem to be trying to make up their minds as to which model to pursue. One school of thought anticipates the Moore’s Law engines of volume and scale will drive more power into the hands of end users, placing control at the network’s edge. The products of this dynamic include the DVR, Blu-ray players and Google TV, devices that attach to networks and allow for brew-your-own entertainment experiences that are nothing like the point-to-multipoint symmetry of old-school broadcast television.

But as compelling as that vision may be, it turns out that things don’t exactly move in orderly alignment. Even as multichannel video providers feed the nation’s DVR appetite – more than 40 percent of U.S. homes now have one, says researcher Bruce Leichtman – an opposing “network DVR” movement wants to stuff the DVR’s capabilities back up through the network pipe and into a centralized server that does the same thing, but without the need for local storage.

Also, telecommunications providers lately are warming to the notion of so-called “cloud” networks that take the jobs of data and application storage away from local machines and give it to a big server somewhere in the beyond. Both Verizon Communications and Time Warner Cable recently made significant acquisitions in the cloud category for purposes of satisfying rising demand among enterprise customers.

So which direction is the train headed? Probably both. For investors, service providers and customers, the trick is going to be placing the right bets at the right time. Certain applications, content and capabilities will favor edge intelligence. Others will profit from centralized storage and distribution.

But as Kamen’s inventive idea for quenching the undeveloped world’s thirst suggests, relying solely on the models of the past as predictors of the future can leave even the best-reasoned business plans all wet.