The future of computers ain't what it used to be

Back in a previous century — when I was a doctoral student and aspiring academic — I met some interesting researchers who were then trying to contradict (or at least temper) some of the wild claims made about the first mover advantage. Anyone knows the computer industry knows that IBM didn’t invent the mainframe, Sun the workstation or Apple† the PC, but all nonetheless became market leaders. About the only example of a long-term success as a first mover was DEC, and that probably had more to do with Ken Olsen than first mover advantage per se.

It turns out the story is similar in mobile computing devices (aka smartphones), as I’ve shown in my research. Motorola invented the handheld cellphone and now they’re owned by a hosted software company. Symbian invented the smartphone category but are now defunct. Apple was a latecomer to smartphones but transformed the category and now makes most of the profits.

Among the academics I met in the 1990s trying to debunk the first mover fable were Peter Golder, Marvin Lieberman and Steven Schnaars. It took a certain amount courage to go against the conventional wisdom, although there were also handsome rewards — for each, their most-cited work (the coin of our academic realm) was that challenging first mover advantage.

Like me, Schnaars has been interested in the historical view of innovation. I was thus not completely surprised to discover this week his 2004 paper on the history of forecasting the computer industry, written with one of his CUNY doctoral students. Here is the official citation and abstract of the paper:

Technology in Society
Volume 26, Issue 1, January 2004, Pages 1–16

Predicting the market evolution of computers: was the revolution really unforeseen
Steven P. Schnaars, Sergio Carvalho

Abstract
This is a study of how people viewed the future prospects for computers from their inception in the mid-1940s to their establishment as an unequivocal market success in 1964. Based on a reading of the published record it compares two hypotheses, one that holds that no one foresaw the tremendous potential of computers—an error of under guessing—the other that holds that most market forecasts are far too optimistic, a error of over guessing. It concludes that only before 1950 did experts fail to foresee the ways in which computers would change our lives. After that date, expectations soared saved only by the fact the sales likewise soared.

Keywords
Computers; Forecasting; Prediction; Market evolution

(If you don't have access to academic journals, Google Scholar will help you find an unofficial copy).

In examining expectations of the future of computing from the 1940s until 1964 (the IBM S/360), Schnaars and Carvalho identify three phases (see Figure 1): a pessimistic phase (the 40s), a wildly optimistic phase (the 50s) and then a realistic phase (1958 onward).

The first and last paragraph of this nicely written article sum up the story:

In 1943, just before the creation of ENIAC (Electronic Numerical Integrator and Computer), considered by most experts to be the world’s first general purpose, digital computer, Thomas Watson Senior, Chairman of the IBM corporation, was quoted as saying: ‘I think there is a world market for about five computers.’
…
In closing, the Thomas Watson Sr quote with which this paper started has served to provide a misleading vision of how experts viewed the future of computers. His son Thomas Watson Jr. did much better, albeit 20 years later when he said: ‘there doesn’t seem to be any real limit to the growth of the computer indus- try’ [24]. In the more than 30 years that have passed since he made that prescient statement, the computer industry has soared.

One point they make obliquely: wildly optimistic forecasts are crazy for most product categories, but once in a while they’re true. AT&T wildly overestimated the demand for PicturePhone (remember the movie 2001?) and wildly underestimated the demand for cellphones — but that’s a story for another day.

† NB: If you want to measure PC success by market share rather than profitability, substitute HP for Apple — but the point is the same.