How Dynamos Can Predict The Rise Of AI


The Podcast

For a while now I’ve been working my way backwards through the fantastic podcast 50 Things That Made The Modern Economy. In it, Tim Harford explores an enormous range of topics and does a remarkable job of summarising their origins and impacts on the world in under ten minutes.

The episode I was listening to that inspired this post was on the dynamo. Tim explains how the dynamo was one of the defining factors that allowed for a total upheaval of manufacturing. The dynamo paved the way for the first commercially successful electric motors around the 1870s. This transformed the manufacturing industry in ways that would have been unimaginable at the time.

Re-Arranged Factories, Re-Arranged Business

In the era of steam, factories had to be arranged in the order that best suited the steam engine drivetrain. In the era of electricity, power is transmitted through wires that can be arranged in whichever way the owner pleases. This allows the factory to be arranged for the benefit of the production line.

In other words, the limiting factors that governed rates of production shifted from being limited by steam engine output to being limited by the workers on the production line. To take full advantage of the new technology, no longer could a successful business focus solely on the quality of its engine; a successful business was one that could rearrange itself and adapt to the new limiting factor: people.

Businesses struggled to ditch the old ways and move to new styles of working for many and varied reasons. However, around the 1920s, America and other parts of Western Civilisation began to see an enormous increase in productivity. The economic historian Paul David argues that this was not to do with any technological jump at the time, rather, it was businesses successfully adopting the large organisational changes. The ~50 year lag time had finally brought with it the benefits extolled by those who could see into the future of industry.

A Modern Equivalence

This draws stark parallels with the transformation that has occurred as a result of the digital age. Around the year 2000, we have again began to see an uptake in productivity. Interestingly enough, the first computers began to take hold roughly fifty years prior to this… In a fantastic paper by Erik Brynjolfsson and Lorin Hitt of MIT, there are numerous examples of firms that have re-organised to take advantage of new ways of working. They outline this change process to involve two distinct stages: the enablement of complimentary working processes and the product improvements that result from those processes.

This ties in very well with those of us who have seen both the successes and failures of businesses in embracing the digital era. For those involved in such things, how many times can you count the phrase “digital at the center” or “agile transformation” in your recent conversations about business strategy? These are the new forms of organisational change that unlock the fabled productivity benefits.

The Next Cycle

As this cycle continues, we can ask ourselves “What is the next revolutionary technology?” Most of the top answers these days are likely to revolve around Machine Learning, AI and the IoT. Some dominating market presences have figured out how to organise their business around these technologies already, however, widespread adoption seems a long way away.

As we have seen from the past, envisioning the success of everyday business in these areas can be imagined in the context of what organisational changes it would take for companies to begin to take full advantage of new work processes and resulting product improvements. Key processes that I can see include:

  1. the acquisition of good quality data
  2. the interpretation of that data using complex models
  3. the conversion and justification of that interpretation into trusted information
  4. the enablement of business to respond to that information

None of these sound particularly easy right now! However, having the awareness of the significance of these areas is certainly a benefit in helping to prioritise work for the next generation.

© neverstew 2022
GitHub logo