The ultimate stage is reached through a systematic maturing in ever more reliable, simple, and efficient steps; the more complex a system, the further it is from its ideal and final stage. When final, the system has no components any longer, just pure function. Thus, for example, the ideal lawn mower would be no such thing – it would be eliminated – [instead we would have] genetically modified grass that stops growing at the ideal height, with Wimbledon density.”
Wouldn’t indeed innovation benefit from some iron laws as to what might be expected next? Future studies, or futurology, or futuristics, have sometimes attempted to presage technological (and also other) trajectories. A series of steps in the development of a function, the velocity of fighter aircraft, the light intensity of lamps and light sources more generally, might describe a smooth curve, the envelope of a series of discrete evolutionary (or even revolutionary) steps and technological transitions, adding up to, for example, an exponential curve, possibly modified into a biological transition, thus an eventual leveling off. The most well-known such trajectory is probably the one for what may be termed computer power, encapsulated in Moore’s Law.
In the mid-1970’s, William J Abernathy and James M Utterback proposed a model for how a technology evolves over time. Early on, there is a fluid phase, characterized by little or no standardization, lots of new actors and the concomitant experimentation. As the technology gains in importance and investments accrue, money and competence-wise, technology stabilizes to eventually converge into a ‘dominant design’ (Utterback: Mastering the Dynamics of Innovation, Harvard Business School Press 1994). Chances are that this dominant design crowds out all others though W Brian Arthur has demonstrated that a ‘minority one’ may co-exist. Finally, there is the third stage, the specific phase or maturity, with few and large actors, standardized products, innovation having switched from product to process innovation before that, too, peters out into refinements.
Arthur is closely associated with complexity mathematics and the economics of increasing returns. More recently he has turned to reflecting over the nature of technology, how technology develops. The first impulse would be to think of it as something evolving through some kind of Darwinian evolution, however, after deeper considerations, Arthur says ‘not so’ but for incremental steps. It is rather, he asserts, a process of bootstrapping, existing technology being refined, built upon, possibly boosted by new knowledge from, say, users or scientific breakthroughs – new combinations. Bootstrapping equates with positive feedback, thus Arthur’s old increasing returns economics.
The nature of technological progress also explains something that has been observed before (particularly by economist Paul David): a kind of incubating before the true potentials of a new technology are coming to fruition. The classic example is the transition from mechanical systems and steam power to electricity and hosts of electrical motors substituting one central energy source. Despite arguments and forecasts nothing much seemed to happen in the years after, say, 1880. It took until around 1920 for the long-awaited transition to take place. This, we may conclude in retrospect, was because of the need for the associated structural changes to be instituted – factory buildings pure and simple, knowledge and competencies to evolve, organizational adaptation to happen. Brian Arthur details some further examples, from history and out of what is yet to come. – Renowned French historian Fernand Braudel is famous for his analysis of history as progressing at three different paces, for different mechanisms (geographical time or la longue durée; social and cultural change; and events horizon). He has been criticized for not offering the links between the three. Perchance Arthur’s description opens a way?
Our piece under this Misc. heading deals with that most visible indicator (sometimes false, misleading, or just insufficient) of innovation: patents (cf my September 28, 2010 piece). FreePatentsOnline is a patent search engine and it certainly offers food for thought. The map of patenting at-this-moment reveals its US focus and base.
An economic corollary to Moore’s Law implies that this very law might become self-destroying because of the immense costs for building a next-generation factory (“fab”) to achieve the precision necessary – following Moore’s Second Law or Rock’s Law about exponentially increasing costs. For other information technologies, advances are also fast, exemplified by Kryder’s Law for hard disk storage costs: another envelope for innovation and its results. Likewise, Butter’s Law of Photonics parallels Moore’s Law though here speed is even higher, the amount of data carried by an optical fiber doubling every nine months. Thus, the cost for transmitting a bit over an optical network decreases by half every nine months. And Nielsen’s Law claims that bandwidth available to users is increasing by half annually.