Growth & Robots

What is shown here is a graph of per-capita GDP growth, starting in the year 1300 and projecting to 2100.  Neat, huh?  This is projecting the possibility that for my kids and their kids, they will be living in a static time where things pretty much stay the same generation to generation.  No problem, if by then we finally all have flying cars and universal health care – not so great if we don’t.

Paul Krugman has lately been writing about this, and another really interesting and related idea, the role of robots in future growth.  At IBM where I work I’m a member of the IBM Academy of Technology, which does a yearly internal survey of members on business, technology and social trends for the coming year.  In answering the survey Krugman’s recent observations were much on my mind – however the predictions play out I think these are critical challenges for the next 20 years.

Here’s the substance of the two ideas.  On growth, the seed is a paper by Robert J. Gordon, entitled Is US economic growth over? Faltering innovation confronts the six.  Gordon observes that the Western world has had three periods of growth: The first started in the 18th century, lasted roughly 100 years and introduced steam power and railroads – this is what is typically termed the Industrial Revolution.  The second was a technological revolution, running from around 1870 to the Great Depression; this phase brought electricity, internal combustion, chemicals, petroleum and control of infectious disease.  The third phase, shortest and most recent, started in the 1960s and is all based on information technology and communications.

This arc of technological progress is familiar to everyone in the developed world and it’s central to our implicit belief that things will always get better.  The narrative goes like this: My grandparents as kids had neither electricity nor indoor plumbing; then my parents had both but no information technology; and now we have online banking, digital entertainment and hybrid automobiles.  Can the technological singularity not be fast approaching?

Gordon’s thesis however is that the growth brought by the IT revolution is far less fundamental than the growth of phase #2.  Phase #2 included numerous major innovations that created new economic baselines but were not repeatable.  For example air conditioning was a great advance in that it allowed people to live and be productive in times and places they could not before; but now that we have those productivity gains AC isn’t adding any more new benefit.

So, growth may not be a permanent state: Many of our major innovations were 1-time things, with no expectation their performances will be repeated.  And our current phase of growth has run its course in terms of productivity and now focuses on areas like entertainment.  In his paper Gordon proposes this thought experiment:

A thought experiment helps to illustrate the fundamental importance of the inventions of IR2 compared to the subset of IR3 inventions that have occurred since 2002. You are required to make a choice between option A and option B. With option A you are allowed to keep 2002 electronic technology, including your Windows 98 laptop accessing Amazon, and you can keep running water and indoor toilets; but you can’t use anything invented since 2002.

Option B is that you get everything invented in the past decade right up to Facebook, Twitter, and the iPad, but you have to give up running water and indoor toilets. You have to haul the water into your dwelling and carry out the waste. Even at 3am on a rainy night, your only toilet option is a wet and perhaps muddy walk to the outhouse.

Which option do you choose?

Not much of a choice, is it?

It’s hard not to be suspicious of the sticking-power of today’s IT driven growth, especially when you consider there are probably 1 billion pictures of cats on the internet.  Anyway what Gordon is saying is that phase #3 has a great many incremental improvements, but in the larger scheme of things few game-changing innovations.  Thus, phase 3 brought us a flare of intense growth but is diminishing as the world quickly becomes saturated with PCs, phones and software.

Now let’s turn to the robots notion.  Krugman’s first post on this was titled (in a nod to Asimov, I suspect) Rise of the Robots.  In that post he observes that labor’s portion of GDP – i.e., wages and direct payments to employees compared to total production – has been dropping steadily since the early 1970s; this is after staying mostly steady through the 40s, 50s and 60s.  In a way that is to be expected – since we are becoming more productive, each $1 paid in wages should be producing more and more goods (although I suspect socially and politically there’s more going on here than just productivity improvement).

The question is, where will it end?  A telling event is Asian tech manufacturer Foxconn’s announcement to deploy 1 million robots.  Foxconn currently employs 1 million workers.  While the company states the purpose of the robots is to enable “advanced manufacturing”, no one doubts that a great many workers will be displaced.  And this is in a region where wages are amongst the lowest in the world.  On the other hand, automation may enable the return of manufacturing to the developed world, just as Apple has stated it is investigating.

Regardless of where it happens this trend in automation could lead to even greater economic inequality.  More and more profit will go to those who own the means of production – i.e. the robots – and less and less to people who work.  Machines displacing workers is nothing new – that is how the term sabotage was invented after all – but in the recent past we have had growth to reduce the impact: Workers got less of the pie, but the pie was growing really quickly.  Now pie-growth is slowing and may even stop.

After all this I have no big conclusion – certainly not today.  Questions on my mind:

• Are there real, move-the-needle innovations yet to come from the IT revolution?  Or will we be just using algorithms and machine intelligence to make our existing stuff incrementally smarter?
• What kind of opportunities are there for individuals or small teams in the upcoming automation revolution?
• Do we need to re-examine our ideas of growth?  Maybe Stiglitz is right, that we need a better GDP, that includes externalities and quality of life and not just raw production.

Like I said, don’t know.  Maybe this is the future: