Explore a better way to work – one that promises more calm, clarity, and creativity.

On the Dynamo and Email

In an article about remote work that I wrote for the New Yorker last year, I pointed to an underground classic research paper titled “The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox.” It was written by a Stanford economist named Paul David, and published in the American Economic Review in 1989.

In the article, David performs a close study of the adoption of electric dynamos in factories at the turn of the twentieth century. With the benefit of hindsight, it’s obvious that the right way to leverage electric power in factories is to put a small individual motor on each piece of equipment. As David points out, however, it took decades after the introduction of practical electrical generation before this obvious shift finally occurred.

As I summarized:

“[A]t the turn of the century, most factories were powered by massive central steam engines. The engines turned overhead shafts, which were connected by an intricate array of belts and pulleys to close-packed machinery. When electric motors were first introduced, factory owners tried to integrate them into their existing setups; often, they’d simply replace the hulking steam engine with a giant electric dynamo. This introduced some conveniences—no one had to shovel coal—but also created complexities. It was hard to keep all the electrical components working; many factory owners opted to stay with steam.”

I think this example is useful in orienting our thinking about the technological disruptions afflicting commerce in our current century.

The arrival of digital networks in the office is similar to the arrival of electrical power to factories. In both cases, the innovation is accompanied by a confident intuition that the very nature of how we work is poised for an irreversible shift toward a new, better configuration.

But as David’s paper underscores, these shifts can be slow to unfold. New technology is often a necessary but not sufficient precondition for revolution. The other ingredient is the uneven, frustrating work of changing the definition of “work” to accompany the new tool.

At the turn of the century, factories had all of the components needed to deploy the much more efficient individual motor approach to manufacturing. But it still took decades for them to actually make the shift away from hulking central power plants and overhead spinning shafts.

Today we almost certainly have all the technological tools needed to push knowledge work into its next productive phase shift. But we remain in the moment mired to instead simply moving unstructured interactions into email threads and Zoom meetings; the Digital Age equivalent of hooking up a new electric motor to the old belt drive system.

The solution is not to look to Silicon Valley and ask for better innovations. The needed technological innovations have already arrived in the form of Ethernet, email, and interactive HTML. Where we need to turn our attention is back toward our own proverbial factories and ask if there’s a better way to run our machines.

12 thoughts on “On the Dynamo and Email”

  1. This article has my head spinning and considering how healthcare is structured to intake new patients. Systems with less resources continue to rely on paper methods or a combination of paper and online records. I’m curious it this applies in anyway.

  2. The argument you put out is incomplete and an oversimplification.
    First factories had to be by the river to take advantage of water power.
    The steam engine was a revolution, because the factory did not have to be by the river. You had to have water, sure, but the water was not the power source.
    How would you retrofit and existing factory with electric motors? Throw out all of the existing equipment and install new tooling with individual electric motors? Because the factory owners did not do that, something was wrong. No you hook up a motor to the existing tools and remove a few belts. Repeat as required.
    Maybe the digital mistake was believing email would remove a few typewriters. It is different than a paper memo, a whole new animal.

    • The point is more general – it takes us some time, and a few wrong moves, to get it right with new technology.

      I worry that it’s been 20+ years of network/internet usage in the workplace, and yet we seem to be doing worse every year – as Cal has effectively detailed. Moving electronic hardware and software technology around is much easier than multi-ton engines, belts, etc., so we should be doing better with the new stuff by now.

    • I wonder from an environmental point of view which is better?
      A lot of industrial works are now air powered so they still use this model which makes you think it must be cheaper to run a single large compressor than lots of smaller electric motors.
      Also email is dying, IM (Teams, Zoom, Lynx, Skype etc) are taking over and they are a different animal again.
      I think the big problem which is write large in computing is the wave of the future and is it really. You can see it in electric cars as no one really knows what will work going forward and at some point everyone will get on the same page and it may not be the right page.

  3. The attention economy combined with the anxiety/pressure to “produce” something leads to situation & the overuse of email/meetings.
    Maybe there’s a shift to be done regarding the metrics used to assess work efficiency.
    Can’t be the same for every sector…

  4. >The solution is not to look to Silicon Valley and ask for better innovations. . . . we need to turn our attention is back toward our own proverbial factories and ask if there’s a better way to run our machines.

    that’s exactly what the tech industry is delivering. New machines and better ways to manage a project, do recruiting (ATS), engage with customers, produce videos, track expenses, manage your team. That’s just a few and there are countless products in each category I mention.

  5. That’s exactly what the tech industry is delivering. New machines and better ways to manage a project, do recruiting (ATS), engage with customers, produce videos, track expenses, manage your team. That’s just a few and there are countless products in each category.

    Tech products for business are core technologies, they are sophisticated solutions to complex problems created by extremely talented and customer-focused teams who are striving to deliver real value.

    • Unfortunately the practical outcome is often less than what the IT folks sell us. I remember an expense report software that my company started using. The accountants loved it–it made it far easier to deal with the myriad of regulatory issues that a multinational company with public and private clients has to address. The people using the expense reports rebelled. What used to take us 20 minutes could take as long as 3 hours due to the program crashing repeatedly (it had about a 30% chance of crashing every time you made a change). Things were counter-intuitive, resulting in a huge error rate and a huge uptick in rejected expenses. It was so bad the company went back to the vender to have the program fixed.

      That isn’t a unique situation; I’d say about half our updated software has issues like that. Most aren’t on that scale, but it’s the same issues. Ultimately what occurs is that individual users build their own methods for tracking the data and transfer it to the company-mandated programs. This doubles the effort and eliminates any benefits from the program. It’s like going out to dinner, but being told you have to cook your own food and bring it to the restaurant. Why bother going out to eat at that point? It’s just an added step with very little value added–you still have to do the work yourself.

      To be clear: I’m not saying all new systems are like that. But a surprising number are. Most companies would probably do better to find a program that suits its needs and stick to it, rather than changing programs regularly.

  6. David’s historical comparison is interesting. It reminds me of a comparison a friend of mine made of the effects of the Gin Craze on 18th century society with the effects of social media on modern society. In short, suddenly everyone could get drunk quite cheaply, so they did, even feeding gin to babies to keep them quiet! The effects were…predictably…disastrous.

  7. The findings in Paul David’s research paper reminded me of a similar discussion in Brynjolfsson and McAfee’s, The Second Machine Age. Sure enough, they cite to the same paper. Well done!

  8. I agree with your assertion that we’re not fully leveraging the tools at our disposal in the realm of knowledge work. It’s clear that we’re on the cusp of an entirely new way of working, but perhaps our collective mindset is yet to catch up with the technological revolution.


Leave a Comment