Why Automation Fears Are Misplaced

To my utter astonishment, every nook and corner of the internet seems to be suddenly swarmed with hand-wringing reports sounding the alarm about how “automation”—robotics, AI and machine learning—will replace humans and otherwise cause job loss, or how automation will adversely affect or impact businesses and/or society.

These adverse automation effect admonitions have even emanated from the U.S. government; last December, the Obama White House released “Artificial Intelligence, Automation, and the Economy.” But this cautiously optimistic presidential report is an outlier; most of the “automation” stories in circulation consist largely of the click-bait sky-is-falling variety.

Are our memories that ridiculously short? Fears of machines taking the jobs of humans or trying to figure out how to capitalize or survive automation date back more than a century, to the dawn of the industrial revolution. This hand-wringing is almost always accompanied by either dystopian disaster or by utopia fantasies, with machines doing all the work while humans lounge around with the time to reach their intellectual potential. H.G. Wells actually portrayed both futures in “The Time Machine” with lounging Eloi and evil machine-centric Morlocks. This was in 1895.

Wells isn’t alone. In the popular imagination, automation advancements are always seen as evil. Being slaves to machines is the primary theme of the dystopic Metropolis (1927), and the dehumanization of the assembly line was brilliantly satirized in Charlie Chaplin’s 1936 classic, Modern Times. In more relative modern times, “artificial intelligence” was seen as a threat to human office workers 60 years ago in the Tracy-Hepburn comedy Desk Set (1957). And Wells’ The Time Machine has been made into movies multiple times, delivering the same automation-is-bad conclusion over multiple generations.

Over-wrought Worries?

As usual, reality runs counter to fictionalized fears. Never have extreme machines-replacing-human fears been realized, and they are unlikely to be realized now or in the immediate future.

With the obvious exception of the damnable cotton gin, history is replete with businesses and society successfully benefiting from automation advancements, even considering short term and limited job loss. For instance, the increased efficiency of the McCormick reaper was seen as a threat to individual farmers. The telephone wrecked the human-dependent telegram economy and, later, automatic switching ended the era of human telephone operators. Automated electric street lighting eliminated manual street lighters. The automobile displaced blacksmiths, carriage makers, stables and the rest of the horse livery business. Henry Ford’s assembly line was seen as a threat to the individual artisan.

Why is automation almost always a good thing? For starters, the number of human jobs actually multiply when new automation industries take hold. Someone has to invent and design the new gear, new factories have to be built, workers are needed to build the new gear and replace infrastructure, etc. Then there are the complex supply-chain pieces that have to be created, built and maintained to support the new automation.

What’s ironic is the medium through which all this automation effect hand-wringing is taking place is the internet, itself an automated replacement for more human-centric manual mass communication media. And the internet information economy certainly has created more jobs than have been lost in the transition.

Real Automation Dangers

While modern automation is unlikely to bring either utopia or dystopia, there are nonetheless dangers. Not from potential job loss or the inability of companies to adapt or exploit automation (which creates internal and consulting jobs in and of itself), but from the ever-widening digital divide.

What’s different in our current era is not automation itself but the speed of change when compared to previous eras. Automation advancements take time to filter down as infrastructures to support the new automation realities are built. Cumulative advancements in multiple disciplines—including processing power, broadband speeds, AI, robotics, materials research, machine learning, even power-generating and ecological technologies—as well as local, national and global economic, political and security considerations, are accelerating faster than ever. This acceleration is making it difficult for even developed-world entities to absorb automation advancements, which is what’s triggering the current automation sturm und drang. There is no guarantee that what is adapted today will still be relevant even a year from now.

But as more than a century of automation disruptions have illustrated, society is almost always the long-term winner, with success defined by those who innovate, embrace and exploit automation, not those who avoid, reject or fear it.