Automation, Jobs, and the ROI Water Level

Scott Francis
Next Post
Previous Post

You can’t focus as much as we do on process innovation and improvement without also thinking about what can be automated and what belongs in the domain of human judgment.  In BPM is Everywhere, we argued that in spite of all the progress in technology, business is still about people, and we shouldn’t abandon the basics.

But another lens through which this same issue plays out is the tension between machines and jobs.  At what point will machines do jobs that currently humans do.  I think Adam Elkus has the most succinct write-up of that tension and its inevitably unsatisfactory answers. We focus on what jobs “can’t be automated,” but he points out:

“In reality we have very little idea what kind of skills will be in favor in this century”

I think this is just one of those statements that cuts right to the chase.  We don’t have the ability to predict:

  • How technology will evolve,
  • How our needs will change
  • How constraints will change (cost of energy, interest rates, education, economic growth, etc.)
  • How salaries for specific jobs will fare (supply and demand is notoriously hard to predict).

Beyond that, it is even harder to predict how man and machine will co-evolve. 

I like the author’s treatment of measurement error between qualitative and quantitative measures.  Our computers are dramatically faster, but that doesn’t mean that we experience a similarly better outcome when we’re doing word processing.  The washing machine that can’t go to the moon is a great cautionary tale that accurate measurements don’t necessarily lead to accurate or obvious conclusions.  We have to remember what they’re actually measuring.

So what are the other problems with defining the limits of automation, and the impact on jobs?  Well, effectively it comes down to the fact that the goalposts are indeterminate, movable, and relative to our current context.  Beyond that, we’re just not that good at predicting the future, and as the author states, we’d be making two wagers

  • “I can successfully predict the course of future basic and applied research and development in the underlying science and technology.”
  • “I can successfully predict how science and technology will change society and how society will change the underlying science and technology.”

As I’ve been working on “automating” for my whole software career (in a sense, isn’t all software development “automating” in one sense or another?), these are really interesting perspectives on the subject.  That tipping point always seems to be just ahead of us, over which jobs will vaporize in an onslaught of automation.  But so far, the number of jobs continues to increase, the automation is probably better referred to as augmentation, and we leverage technology and automation to set higher standards and bars for future performance.  In short, the goal posts keep moving. 

At BP3 we try to focus on making people more effective through better processes. Putting people and customers at the center of that process innovation and improvement has been instrumental in delivering great solutions alongside our customers. But it is also part of how you make peace with reducing work loads – the goal is to enable productivity and people, to make each person more valuable to the business.

I’ve described the “work that can economically be done” in the world as a pyramid, submerged mostly in water.  The water line represents being just above break-even on ROI.  Anything below the water line isn’t worth doing, because it costs more than the benefit.  Anything above the water line is worth doing.  If we can make software better, or make people more productive, then in effect we’re lowering the water level, exposing more work that is worth doing  – more of the pyramid.  How you feel about this phenomenon depends largely upon what you think the shape of the pyramid is:

Before After

Right Side Up.  This means that the amount of work worth doing increases geometrically with respect to a lowered water line.

Before After

Cylindrical or Narrow. Work available only increases linearly, not geometrically.

Before After

Upside Down. The increase in work is diminishing at a geometric rate at some level.

Before After

Melting pyramid. The work available for ROI doesn’t increase – the pyramid gets smaller at a rate that equals or exceeds the lowering of the water line… And if it gets smaller at a faster rate than there’s less work available. If it gets smaller at a slower rate then there’s more work available but maybe not as much as you’d think.

All potentially valid, but reflecting different world views on the impact of software on our work and lives.  If, like us, you empirically experience the original pyramid diagram at the top, then looking for ways to lower the water line and expose more fertile ROI ground is a worthy goal.  

 

 

 

Related Posts

  • Shravan Singh

    Nice article !! With the recent technological advances in AI and Cognitive Computing its very difficult to predict where automation will lead us down the line. As Ray Kurzweil points out that humans are biased to think linearly but technology grows exponentially !!

    • exactly. it is hard to estimate the impact of technology on work. Did email remove work or create it? Hard to say explicitly but we can definitely judge that email made a lot of communication economically viable that was previously not worth having.