With these eloquent words, Greek orator Cicero reminded us of the importance of grounding our view of the future in lessons from the past. As we ponder the future of the Internet, we can learn a lot about where the Internet is headed from the evolution of other networks. A look at the history of the telegraph, telephone, utilities, railroads, broadcasting and highways reveals fascinating patterns that have played out time and again. Here are some of those lessons from history.

Gloom Leads to Boom

And boom leads to bust. The human mind tends to extrapolate the past linearly into the future. But network adoption follows an S-shaped curve. Our linear thinking creates a systematic forecasting bias. In the early stages of network evolution, forecasts tend to be pessimistic because the potential of the network is not well understood. However, as the network starts to grow, predictions become overly enthusiastic. This, in turn, sets up a bust when the growth starts to slow.

In 1844, Samuel Morse could not convince the U.S. Congress to fund long-distance telegraph lines, as lawmakers failed to realize the commercial potential of the telegraph. By the time the first trans-Atlantic cable was laid in 1858, the pendulum had swung to the other extreme. The media and industry experts were declaring that the telegraph would embrace the whole world, bringing with it peace and prosperity. Similarly, in 1877 Western Union refused to purchase the Bell patents, seeing no future in them; it later tried to set up a competing company, only to be challenged in the courts and forced out of business.

The Internet and the wireless industry show a similar pattern. In 1996, when the Internet was in its early days, Morgan Stanley issued a report that predicted the worldwide Internet user population at 150 million by the year 2000. The actual number turned out to be 400 million. Similarly, in the cellular industry, experts in 1990 predicted that the U.S. cellular subscriber base might reach 25 million by 2000. They were off by a factor of four, with the actual number being close to 109 million.

Now, the pendulum seems to be swinging dutifully in the opposite direction. Consider the forecasts of U.S. Internet advertising spending made by eMarketer. In 1998, eMarketer predicted that Internet advertising would reach $8 billion?a number it revised to $15 billion in a forecast made during December 2000. Now, these forecasts look wildly optimistic. In 2001, Internet advertising grew only to $7.6 billion, and eMarketer now predicts a 2002 number closer to $10 billion. Meanwhile, in the wireless industry, after years of underprediction, forecasts of wireless users now range from optimistic to euphoric. If history is any guide, the real numbers will be far lower than the most pessimistic of these forecasts.