Friday, January 18, 2019

Is 5G a Paradigm Shift?

Some observers use the term “paradigm shift” to describe 5G. Without getting overly picky, “paradigm shift” is a term of art, coined in 1962 by Professor Thomas Kuhn to explain the way science advances. In a nutshell, he argued that science progresses in a non-linear way.

The given consensus begins to encounter greater and greater anomalies, until finally the old consensus abruptly breaks, and a new interpretation arises. Precisely how that applies to industries and business models is never straightforward, especially since marketing grandiosity gets in the way.

But consider one way of looking at scientific paradigm shifts:
  • The transition in cosmology from a Ptolemaic cosmology to a Copernican one.
  • The transition in mechanics from Aristotelian mechanics to classical mechanics (motion).
  • The acceptance of the theory of biogenesis, that all life comes from life, as opposed to the theory of spontaneous generation.
  • The transition between the Maxwellian Electromagnetic worldview and the Einsteinian Relativistic worldview.
  • The transition between the worldview of Newtonian physics and the Einsteinian Relativistic worldview.
  • The development of quantum mechanics, which replaced classical mechanics at microscopic scales.
  • The acceptance of Lavoisir's theory of chemical reactions and combustion in place of phlogiston theory known as the Chemical Revolution.
  • The acceptance of Mendelian inheritance, as opposed to pangenesis in the early 20th century

The notion here is way beyond “big change.” Such paradigm shifts imply quantum changes (fundamental changes of state) that are “sudden, dramatic, and enduring.” Much of the emphasis here is on sudden.

It might help to look back at industry changes that could qualify as a paradigm shift, when old assumptions about the business model were broken, and were reconstituted in a new way, globally.

It is likely reasonable to require that any proposed paradigm shift affects the whole global industry, not a single industry segment or country.

A bit harder are platform changes within the industry: analog to digital switching; fixed to mobile; copper to optical access; monopoly to competitive regulation; state-owned to private ownership of firms.

Of those important changes, only the regulatory change from monopoly to competition--and government-owned to privatized firms--might universally be considered a paradigm change.

And that is a key facet for understanding when a development really is a “paradigm shift.”

Though quantum change and paradigm change often result from a long, slow period of quantitative changes, the shift of paradigm is quite sudden. As big a change as the change from “copper to optics” and “analog to digital” have been, some might not consider them quantum or paradigm changes.

The shift from fixed to mobile arguably was a paradigm change. The shift of revenue models (voice to internet access; fixed to mobile) might also qualify, but is more debatable.

In the addition to the end of monopoly, and the shift from fixed to mobile, the internet is probably the only other clear candidate most would agree was a paradigm shift.

Many firms now can use the internet to compete directly with telecom services or displace them entirely. At the same time, internet apps and services allow consumers to change their behavior, reducing aggregate demand for telecom versions of communications and entertainment products.

So Is 5G a paradigm shift? Some of us would have to conclude it probably is not. Is 5G a possible “revolution?” Yes. Even if not a paradigm shift, 5G could usher in an era where industry growth shifts from consumer to enterprise revenue sources. Will 5G pose a fundamental challenge to fixed network revenue models, on the older model of mobile substitution? Possibly, in some markets.

Failing that, 5G might still qualify as  “a big change.”

Claims that 5G is a paradigm shift are probably way overblown and hyperbolic.

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...