Innovative technology, jobs and the razor’s edge

According to the Wall Street Journal, the unemployment rate in the U.S. is nearly 8%.  For people under age 25 unemployment is 16%.  In Greece, the employment rate is 26%, double for young adults at 58%;  Spain and Italy have similar unemployment challenges.

What is making jobs in economically developed nations disappear?  Could it be the effect of offshored manufacturing?  Frugal consumer behavior that’s shrinking company bottom lines, hence triggering layoffs?  Or…  the unintended side effects of advanced technology?

The answer is all three.  But let’s look at the unintended side effects of advanced technology.

Technology is a double-edged sword.  One the one hand, technologies that are more efficient than humans improve the quality of human life.  On the other hand, these same technologies destroy jobs.  New technologies improve food production and health care, liberate creative people from the shackles of the middle man,  and increase social mobility.  Yet, as technological advancement accelerates, software programs, machines and robots are becoming better employees than we humans.

We walk a razor’s edge.

Technological innovation is frequently offered as a panacea for what’s wrong with our faltering economy.  It’s a pretty widely accepted notion, at least in mainstream circles in the western world, that technological innovation increases economic growth, which in turn introduces high paying jobs, new industries, and new efficiencies that in turn, begat new industries and high paying jobs.  Here’s a typical belief expressed in a white paper on patent reform written by the U.S. Department of Commerce.  Note the Department’s unblinking certainty, that technological innovation is the pulse of a healthy economy.

“All major strands of economic thought now recognize that technological change is the primary driver of growth.   In fact, modern economic theory holds that without technological innovation, accumulation of wealth could not be sustained and per capita growth would trend to zero.”[1]

When new technologies replace human workers, economists call the demise of industries and resulting job-loss “creative destruction,” a term that hints at an underlying belief, that in the long term, innovative technology has a regenerative effect.  Economists point out that although Expedia, the internet and airline databases made travel agents obsolete, these new technologies kicked off a cycle of creative destruction.  In exchange for the short-term loss of some jobs, these technologies created an entirely new type of tourism industry and made travel cheaper and more convenient.

Like a slow-moving traditional travel agency that processes paper airline tickets, the theory of creative destruction holds that when a business fails, it fails because it’s inefficient.  Or, in other words, a technologically behind company can’t compete with more efficient companies that have embraced technologically, hence have become more efficient.  The theory of creative destruction holds that prosperity arises from the ashes:  a failed business’s customers will migrate to a more technologically fit company.  The failed business’s former employees will eventually find their feet at a more technologically adept company.

In the book “The Lights in the Tunnel:  Automation, Accelerating Technology and the economy of the future,” author Martin Ford points out how entrenched mainstream economic thought is in the notion of Creative Destruction.  Ford writes,

“the idea that technology will ever truly replace a large fraction of the human workforce and lead to permanent, structural unemployment is, for the majority of economists, almost unthinkable. For mainstream economists, at least in the long run, technological advancement always leads to more prosperity and more jobs. This is seen almost as an economic law. Anyone who challenges this “law of economics” is called a “neo-Luddite.” This is not a compliment.”[2]

What if mainstream economics are wrong and there’s no such thing as Creative Destruction?  What if there’s just a bit of Creative and then mostly…  Destruction?

One of the best recent books on the dance between job loss and new technology is called “Race Against the Machine” by two MIT economists, Erik Brynjolfsson and Andrew McAfee.  This book is wonderfully written — clear, yet rich with data and information.  “Race against The Machine” swims against the tide and makes a compelling case that innovative technology, by removing the need for human workers, is creating a devastating tidal wave of redundant workers.  Or unemployment.

According to Brynjolfsson and McAfee, traditional economic theory attributes high unemployment rates to one of three factors, either “stagnation,” “lack of economic growth,” or “end of work.”  In the “stagnation” explanation of unemployment, the cure for unemployment is to unleash the adoption of new technology. The Stagnation theory assumes that the way to get people back to work is to make better use of new technologies, to remove social or regulatory barriers to its widespread uptake.  In other words, technological advancement = more company profits  = the creation of new, high-value jobs = economic growth.  In my experience, people who work in high tech fields embrace the stagnation theory of unemployment.

Here’s where “The Race Against the Machine” gets interesting:  its authors break rank and argue that innovative technology is not the cure, but the cause for high levels of unemployment.  Brynjolfsson and McAfee argue that the reason that jobs in the U.S. and Europe have been disappearing over the past few decades is not (as popularly assumed) because labor has been offshored to factories located in cheaper, less regulated markets.  In fact, factory automation, not offshoring, has reduced the need for human workers, leading to high unemployment.

We’re in a losing race against machine labor.  Brynjolfsson and McAfee  point out that computing power is improving at an ever-increasing rate and the cost of hardware components is plummeting.  The result is that automated solutions (be it a database, industrial robot, or data mining application) are rapidly becoming more efficient, hence more cost-effective than a band of unpredictable and complicated human employees.

Machines don’t get bored.  They don’t complain.  And they’re a lot more reliable and precise than human workers.

As computing power increases and software becomes more sophisticated, machines are taking the first steps onto what used to be once human-only territory:  being intelligent.  In 2011, Watson, a project from IBM’s research division, won the game show Jeopardy, beating out the show’s former human champion.  Watson created quite a stir; imagine applying Watson’s massive text-scanning deductive power to interpreting a patient’s symptoms.  A computer can process and remember entire libraries worth of arcane medical data; a human doctor cannot.

Watson’s triumph at Jeopardy was reminiscent of the victory of another IBM computer a few decades earlier, Big Blue, that defeated the world chess champion’s Kasparov.  Although Big Blue’s victory caused quite a stir, it makes sense that a powerful, well-designed computer program could beat a human on the chess board.  Chess is a complicated and difficult game that involves a nearly-infinite number of possible outcomes.  However, unlike life, a chess board is a finite environment.

A chess game is bounded.  Though the number of possible chess moves is staggeringly large, a powerful computer is more than capable of rapidly chewing through all the possible permutations of how a game may play out.  Kasparov’s defeat at the “hands” of IBM’s Big Blue was a milestone.  Yet, the way that IBM’s Watson and Big Blue defeated humans was by processing large amounts of data very quickly and then drawing (good) conclusions.  That doesn’t mean that they’re really “intelligent.”

Being intelligent is one of the primary aspects of being human that many of us still confidently assume makes us superior to technology.  How many times have you heard somebody say, “machines will *never* be as good as humans at any task that involves thinking or reacting.”  Another belief is that jobs that involve “people skills” can’t be automated.  Or creative activity can’t be automated.

Creativity, people skills, even decision making — all traits we assure ourselves will remain the sole domain of the human brain — are becoming automated.  For example, ten years ago, most people would refuse to get into the back seat of a car whose driver was a computer.  Today, Google has proven that computer-guided cars that find their way around using a firehose stream of GPS data and complex systems of sensors and interactive software are actually safer drivers than humans.  During Google’s massive road test where automated cars drove around U.S. roads for a few months, the only accident was a human driver rear ended a car being driven by a computer.

In reality, there are very few occupations ultimately exempt from automation.  Blue collar jobs have been automated for decades.  Now computers can do many white collar jobs as well.  Marshall Brain points out that “As CPU chips and memory systems finally reach parity with the human brain, and then surpass it, robots will be able to perform nearly any normal job that a human performs today.”

I’m torn.  It’s one thing to point to repetitive, grueling labor — be it physical or cognitive — and say “technology liberates humans from drudgery.”  But that liberation has a hidden cost:  no labor, no job.  Yet, new technologies , even as they take away jobs, also dramatically improve the quality of our lives.  The phrase, “do it manually” has become a synonym for work that’s error-prone and inefficient.  In addition, by introducing automation into manual jobs, humans are spared work that’s tedious, inconvenient, even dangerous.

The problem is that technological advancement is accelerating, yet the human brain remains essentially the same.

Where does this leave those of us who are fascinated and excited by the steady stream of innovative technologies that flow out of universities, companies, government labs and the garages of DIY tinkerers around the world?  Martin Ford offers a solution, that we must re-adjust our notion of what constitutes “work.”  In some nations, farmers are paid by the government to let their fields lie fallow.  Ford proposes that one solution for massive, technological unemployment would be that the government pay people for engaging in constructive activities, for example, for going to school or for reading books.

Here’s another solution:  freeze technology development in place.  Imagine if somehow the development of new technology were to just stop.  No existing technology would be destroyed.  We would simply continue to live with the technology we have now.  But that seems like a poor cure for technologically-induced unemployment.


[1] Patent Reform, Unleashing Innovation, Promoting Economic Growth & Producing High-Paying Jobs.”  A White Paper from the U.S. Department of Commerce, April 13, 2010.  By Arti Rai, Administrator, Office of External Affairs, USPTO, Stuart Graham, Chief Economist, USPTO, Mark Doms, Chief Economist, Department of Commerce.

[2] The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future.  Martin R. Ford.  Acculant™ Publishing.  2009.

Comments

  1. Bill says

    Here’s what you need to understand. Since 1980, the govt’s national lab operations in all agencies and cabinet depts have been subject to 5-6 major federal technology transfer commercialization laws. A great number of these national labs are GOCO labs (govt owned, contractor operated), and thus paid a performance fee for their operation. In only one lab I’m familiar with (my expertise is DOE labs), that being Los Alamos National Lab (LANL), does the contractor earn a small fee for “economic development” (ED).

    LANL’s ED small fee is also not specifically geared to techXfer commercialization, as much as it is to keep employment & small-biz outsourced routine functions (laundry, bldg maintenance, light construction, landscaping, stores & inventories, etc, etc) afloat… to maintain employment in northern New Mexico, a historically disadvantaged minority region full of poor people. This strategy, originally crafted by NM Republican US Senator Pete Domenici to keep restless Democrats at bay, has not made this region any wealthier or self-sufficient. But unlike all other DOE NatLabs, LANL’s ED part stays in their performance fee contract.

    If you want to tap into the vast treasure trove of DOE national lab technologies, the govt will have to adopt the premise, “what gets measured (and incentivized) gets improved”. This would include specific fee performance awards for NatLab technology commercialization, not as licensing (the easy way to check off that box), but as full, small biz startup creations. Unfortunately, you have an entirely ineffective minority female in charge of this at DOE-HQ, and in her favor I must also state Laboratory management hates the idea of having to be incentivized by the messy process of creating businesses. They are so-so-so above all that crass commercialism.

Leave a Reply

Your email address will not be published.