VON Visions: a fine way to run a railroad but not an internet

Friday, September 07, 2007
Updated: Feb 14, 2009

Copy of the original column from Von Magazine.

Update: After writing this piece I learned that the FCC was modeled on the ICC (Interstate Commerce Commission). It turns out that telecom policy is indeed modeled after transport policy!.

The transition from steam railroads to diesel seemed straightforward enough–you just changed the locomotive and that was that. It seemed like a minor change, but none of the steam engine manufacturers survived the transition. The reason is subtle. Steam engines had personalities. Each locomotive had a particular purpose and place. Diesels are essentially commodities, and the buyers became less dependent upon the relationships with their suppliers.

Telecom is about maintaining relationships with each customer, but with the Internet those relationships have become a limitation rather than a benefit.

Metaphors get stretched beyond their limits as the technologies change, but instead of adjusting our understanding, we sometimes go to extraordinary lengths to adjust reality to match the metaphor. In the case of telecom, the problem is exacerbated by legislation that gives the metaphor veto power over reality. This reaches absurd and costly extremes in the case of the telecom Regulatorium.

The railroad metaphor is fundamental to the entire business model of the industry. Just like the railroads, the telecom industry sells rides (services), and it lays tracks (wires) in order to provide a path for its trains (channels and connections). In this world there is no business model for the automobile. Fortunately Henry Ford didn’t have to depend on the railroad for his roads. People could use the cars to drive around town and into the nearby “country” for a picnic. For local transport, one could also take a trolley or a bus.

Over time the demands of car (and truck) drivers gave rise to investment in better local roads and better country and state roads. In 1919 Lt. Col. Eisenhower was assigned to drive a military convoy across the United States. As President in 1956 he co-sponsored the Interstate Defense Highway System. Fortunately, the military understood the importance of sharing a common infrastructure rather than trying to build its own system.

Today’s Internet is a dramatic demonstration of how much we can do when we can drive our own packets through the network rather than being limited to a service provider’s offering. No matter how benevolent the service provider, the best it can do is give us limited choices, not opportunity. The service providers’ business model requires that we buy their services. They can’t survive letting us find our own paths. We shouldn’t be surprised that they act like the railroad robber barons of the past–that is necessitated by the business model.

If we think about connectivity in terms of the first square mile around us rather than thinking of broadband as the last mile of a delivery system, we can start to question why we need to have a phone (or cable) company forcing us to buy services from it when we have abundant local capacity. How can it define quality for us? And how can an IMS control plane work if it doesn’t control the entire network? QoS and IMS are about managing scarcity and denying us the use of the abundant capacity we already have.

The change from steam to diesel was minor compared to the change from the railroad model of telephony to the Internet model in which we do our own networking and create our own solutions. The Regulatorium was created in 1934 because the industry would not be viable without its protection, and we accepted it because we could only think in terms of railroads. The Internet has shown us that we no longer need to buy rides. A connectivity infrastructure costs less than today’s transport model, and it’s already available–just lying fallow.

Today’s business models depend on scarcity to create value. We are impoverished by a bad metaphor. That’s foolish.

In the 1980s, (the original) AT&T recognized that it had to change and took the lead with its divestiture because it saw it as a chance to focus on its highvalue long-distance business. AT&T was wrong, but it demonstrated that if the industry decides it’s in its interest, it could make major changes.

We’re long overdue for divestiture II. The longer we wait, the more wrenching the adjustment will be. Investors have a chance to make a deal to restructure their industry with the facilities companies acting on behalf of local communities and service companies taking advantage of abundant capacity.

Experience has shown that once the industry loses its protective railroad metaphor, the investors will regret their intransigence. There is little value in owning a tollbooth once everyone has learned to bypass it.

MIT alumnus Bob Frankston was the co-developer (with Dan Bricklin) of the legendary VisiCalc. He later developed Lotus Express. During his 1990s tenure at Microsoft, Frankston initiated the home networking effort, which has made it possible for you to buy a small router and easily connect your home with the Internet. He can be reached at VonMagColumn@bobf.frankston.com.

Bob Frankston Site