Jun 20 2017
Journalist Charles Duhigg has a new book out on the subject of productivity and was being interviewed about it on NPR. I heard him express as a general principle that new technology never increases productivity when first implemented because organizations and individuals use it as a new way of doing exactly what they were doing before. Over time, productivity does increases as users discover new tasks or methods that the technology enables but were beyond the imagination of its early adopters.
We see this pattern clearly when looking back at successful innovations. Looking forward, on the other hand, we can’t assume that every current invention will have future applications that no one can envision today. Our perceptions are biased because we only remember the inventions that have survived the shakeout of confrontation with markets and society.
We can’t generalize from past inventions, but it is true that the successful innovations have to overcome what Marshall McLuhan called the horseless carriage syndrome. Until hearing about this last week, my only exposure to McLuhan had been his cameo appearance in Annie Hall, in which he berates an academic for knowing nothing about his work:
Working in Manufacturing, why should I have paid attention to the work of an academic who philosophized about TV, radio, and print, and died in 1980, when most the media we use today didn’t exist? Perhaps because he had the ability to formulate general principles that are relevant beyond his area of expertise. Perhaps also because his influence endures, as people who have not read his work unknowingly quote him when they say “the medium is the message,” or talk about the “global village.”
Here is a quote from him about Manufacturing from 1964 that, with the possible exception of the choice of examples, could have been found in the latest issue of the Harvard Business Review:
“Thus, with automation, for example, the new patterns of human association tend to eliminate jobs, it is true. That is the negative result. Positively, automation creates roles for people, which is to say depth of involvement in their work and human association that our preceding mechanical technology had destroyed. […] In terms of the ways in which the machine altered our relations to one another and to ourselves, it mattered not in the least whether it turned out cornflakes or Cadillacs.” (Understanding Media, p. 9)
Whether about 1964 or 2017, I would still disagree. Food processing moves faster and at a higher level of automation than car manufacturing, and “human work and association” are quite different in both worlds.
Following are a couple passages introducing the concept of the horseless carriage syndrome:
“The principle that produced the names “wireless” and “horseless carriage” is another instance in that long list that illustrates how every innovation must pass through a primary phase in which the new effect is secured by the old method, amplified or modified by some new feature.” (Understanding Media, p. 323)
“The word ‘wireless,’ still used for radio in Britain, manifests the negative “horseless-carriage” attitude toward a new form. Early wireless was regarded as a form of telegraph, and was not seen even in relation to the telephone. David Sarnoff in 1916 sent a memo to the Director of the American Marconi Company that employed him, advocating the idea of a music box in the home. It was ignored.” (Understanding Media, p. 336)
This theory resonates for me, because of examples I have encountered in Manufacturing, such as the following:
- Machine layouts. 100 years ago, power transmissions to production machines through belts forced machines to be lined up under an overhead shaft. The advent of electrical power removed this constraint and made it possible to lay out machines at crooked angles to one another, as needed to facilitate the flow of materials. Yet you still find “shaftless layouts” of machines lined up in neat rows underneath imaginary power shafts.
- “Smart” part numbers. As discussed in earlier posts, “smart” part numbers are anything but in the age of databases. They may have been helpful with cards and file cabinets; now, they just make the information encoded in them hard to access and reduce databases to “cabinetless files.”
- SPC. Statistical Process Control, as taught today, is still based on 90-year old techniques tailored to the age of mechanical calipers, paper spreadsheets and slide rules. Implemented in software today, their primary use is to demonstrate compliance with external mandates. In mature processes, capable of holding tolerances ten times tighter than required, these techniques are unnecessary; in high-technology processes, they lack the analytical power of more modern approaches. With 21st century data science available, the quality profession is stuck at “paperless SPC.”
The real issue with the horseless carriage syndrome is that it prevents you from reaping the benefits of innovation. A major cause is management forcing the implementers of new technology or methods to justify investments exclusively in terms of their effects on existing practices, dismissing new capabilities t as “intangibles.” The remedy is learning and experimentation in an engineering sandbox, concurrently with continuous improvement in existing operations. The former lets you discover potential uses beyond current practices; the latter, where these uses can make a difference.