In 1956, Frankie Darro changed my life in ways I never could have imagined. I never really met him, nor even knew who he was until years later. Maybe you don’t know who he is either. And maybe that, dear readers, is a metaphor for this post: it’s what you don’t know that should make you wonder why.
Robbie the Robot was a 7 foot tall star in the movie “Forbidden Planet.” He had a large head and in place of an open mouth was a blue light panel that flashed in sync with his mechanized speed. Frankie Darro was a stunt man inside Robbie. Robbie was the most memorable early example of artificial intelligence and seemed to have been programmed with Isaac Asimov’s Three Laws of Robotics, first stated in 1950. Check ’em out.
That movie, and in particular Robbie, had a lasting but barely detectable influence on my life. I got a degree in physics, launched into a decades-long career in computers, and did some (commercial but otherwise not memorable) work with neural networks, the early foundation of today’s AI world.
It wasn’t just Robbie. I remember many TV programs, articles and pundits declaring how automation was a threat to the working middle class. It would obviate their jobs, and worse, obviate their lives. That essentially didn’t happen then, but it has definitely been happening, and I argue in this post, will be an accelerating trend.
Robots Since Robbie
For decades after my childhood encounter with Robbie, robot ideas and concepts seemed modeled on him. But then came “2001:A Space Odyssey” and HAL 9000 and the computer on the Enterprise in “Star Trek” and changed the robot image entirely. I don’t think fans, aside from a devoted few, thought of them as robots. Instead, the idea of big, often yellow, fast acting but rather stationery workers on automobile assembly lines that moved, assembled and welded parts were our new Robbie incarnates. Or those MIT/Stanford/etc. contestants in the DARPA-sponsored self driving competitions, with their multi-camera eyes, Erector Set quality bodies, and dangling wires harking to mad scientists instead of our social companions.
Along comes Deep Blue (chess) and Watson (“Jeopardy”) from IBM, Siri (Apple), Cortana (Microsoft), Alexa (Amazon), and Google Voice/Translator/et al. (Google). We are seeing these robots in our everyday world (except for the IBM mainframe things as of yet) where we use them as sometimes frustrating but ever improving helpmates. And along comes the Tesla and Google and a pack of other self-driving cars. And on the entertainment front, HBO’s “Westworld” is a modern extrapolation of Robbie and his human companions, presented with a hint of moral and commercial conundrum (and violence and sex – hey, it’s HBO).
Robots we don’t likely see or even know about are in the hands of health care providers and scientists and engineers. Just at the horizon are imaging devices that will detect abnormalities even better than trained radiologists, genetically driven test devices that can predict diabetes or heart problems and cancer and on and on, well before a doctor might. Not too far away is the medical tricorder of Star Trek. Engineers use lasers in smart devices for measurements and design, as do scientists. Scientists get great robotics in large telescopes and particle accelerators, but don’t expect these to become pocket-sized anytime soon.
Old robots are getting significant upgrades, too. Industrial, manufacturing ones. Not only the giant assembly line ones but smaller, more dexterous, smarter ones that are also more mobile. They not only make products but work in warehouses to pick parts and deliver them to a loading dock for shipping.
So What’s that Got to Do with Jobs?
Market conditions have always been drivers for and against jobs. At the start of the Industrial Revolution, most workers were in agriculture on small farms or in some form of cottage industry. But as factories were built, there was a migration to cities where the work force was needed. At the same time, agriculture got significant productivity multipliers, so that one farm worker could produce for many instead of for only himself and his family (excuse the gender-specific reference). That trend grew significantly when manufacturing grew to embrace Henry Ford’s production lines, replaceable parts, and all we have come to expect in products today. Post-war (think especially 1950s) manufacturing embraced a growing middle class economy and manufacturing workers became part of it, building products for others while consuming products with their increasing buying power. All was good.
Transportation changes began shaping how manufacturing companies looked at their options. Canals, then railroads, and then trucks were significant in how manufacturing was planned and carried out but was kept pretty much to national confines. Airplanes and ships started to expand first just markets internationally, but then where items were made. It actually became feasible to make parts in one country and ship them to another for assembly into finished products, then entire products themselves were made in one country and shipped to another. As transportation costs, as well as transit times, fell, manufactures saw their significant cost component was labor. And it was far cheaper to find labor outside the US, and manufacturing went where the cheap labor was.
Certainly free trade deals made some overseas migration of US manufacturing jobs easier, but manufacturers would still have chosen cheap labor costs over tariffs; tariffs could be passed along as a price component to consumers, but high labor costs were harder to include in low product prices. The flip side of this is also important: US products in agriculture, technology and other sectors where the US held dominance benefited from the free trade deals. Hold onto these thoughts while we take a side trip in our story.
Changes in Policy Ahead?
Donald Trump’s campaign promises may be fulfilled if reading the tea leaves of his announced appointees is a prediction of the future. These include fining companies who send jobs out of the US, a 35% tariff on (at least automobiles made in Mexico) goods made outside the US by US manufacturing companies, “tearing up” trade deals and re-negotiating them, significantly reduce the federal government’s role in many areas and reduce federal spending by whopping amounts, and repeal and replace the ACA aks Obamacare. He promised that these measures would create <insert your enormous number here> of jobs, increase US GDP growth to 4% annually, bring down unemployment, bring back steel production, coal jobs, and open up US drilling and production of oil.
Let’s examine just a few of those. Steel mills have shut down in so many areas displaced by newer, more efficient mills in China and Japan, for example. Not only would it take significant capital investment to create competitive mills and time to build them, but investors are almost certainly going to look at more lucrative returns on their investment dollars in other ventures. Coal has been dealt its setbacks not so much because of regulation but by cheaper renewable energy like solar and wind, and very cheap and clean natural gas. Coal fired power plants, already an endangered species because of cheaper fuels to run them, may additionally face more pressure from a new generation of nuclear power plants that are safer but still costly and time consuming to build (an likely to need a higher level of government intervention, unless the presumptive DOE Secretary Perry decides nuclear energy should be a completely free market enterprise). And somewhat unnoticed by the media and pundits is that coal (coke) usage also fell when steel mills were shuttered.
Maybe 40% of light trucks sold in the US are made in Mexico. Slapping a tariff on them would probably raise the average price somewhere in the neighborhood of $6,000 per vehicle if not more at 35% on a $17,000 cost. Buyers will struggle to come up with the extra cost, sales will fall, and auto makers will respond by pulling back. It’s not just the workers at their plants in Mexico; it’s also the replacement parts makers, the sales staffs, the body shop employees, and all the auto food chain that will pull back as well.
Substitute any other product made overseas and sold here and you get a similar scenario: smartphones, consumer electronics, clothing, furniture, toys, and even produce like fruits and vegetables, a seemingly endless list of what we have to have at the cheapest prices possible. In other words, put tariffs and duties on imported goods, see prices go up, sales go down, and a whole backlash of things happen. Not to mention the unhappiness of buyers who want all of that stuff.
So Why Did We Lose Jobs in the First Place?
Market forces certainly led to cheaper labor overseas, and it wasn’t only in manufacturing. Call centers, software developers, human resources, tax and accounting, all sorts of services that could be outsourced. Countries like India, China, and other developing countries not only had government policies that fostered their domestic workforce to develop the skills necessary, but they encouraged private investment – often from American companies and investors – to nurture that development as well. So while manufacturing jobs went overseas, so did many white collar jobs. You just didn’t see a lot of HR or tax accounting people holding protest signs during the 2016 campaign.
Which brings up another important issue. Over the decades and centuries, many jobs have been lost. Job categories, that is. Buggy whip makes is a moniker for this but suggests it is a very old and ancient issue. It is not. How about elevator operators, telephone operators, workers who made mimeograph and fax machines, TVs that were not flat screen.
Some, many that is, jobs are lost because the market for what those jobs produced dry up. The jobs are replaced by innovation, new inventions and discovers, economies of scale. You can’t bring those jobs back. And most people would say, why bother?
I started this story talking about robots. Ponder for a few words with me how robots (as a broad class of automation) will affect jobs in the near and even very near future.
Self-driving cars and trucks and drones. They are so close we can touch them. Their impact on jobs will be enormous. It is not just that trucks and cars as we know them will zip along carrying goods without anyone driving them. They will certainly enable very different paradigms of what cars and trucks do, how they are built, and how they function. To illustrate this, consider that Amazon bought a robotics company and has deployed tens of thousands of robot pickers in their vast warehouses. They get order information from computer systems that no doubt have sophisticated algorithms to get items ready first that have the earliest delivery/departure requirements grouped by how they aggregate for shipping. But more to the point, Amazon can now completely change the way it organizes items in its warehouses from what works best for humans (like items grouped together so humans more easily remember where they are) to. say, items often ordered together because computers remember where things are for the robots. Likewise, cars and trucks that don’t have to accommodate human drivers will certainly be different from what we know.
Health care providers may not lose their jobs as fast as truck drivers, but they will require brand new skills and a refocus on how various practices actually work. Teachers will continue to find value in automated teaching aids. Houses now built by crews on their foundations will see the trend for factory-built components installed on site but eventually even those will be done by robot cranes and framers. This is a minute fraction of what can, or more than likely, will be done.
And it’s not just in the US. Foxconn, manufacturers of the iPhone in China, recently completed a phase of robotic automation and reduced their workforce from over 100,000 people to about half of that. And expect to do more of the same. Not only does automation reduce the labor head count and cut out spending on salaries and benefits, it allows new tasks to be done by machines that simply cannot be done by humans. Coal miners have long known this and don’t dig out coal with a shovel and wheelbarrow. Beyond brute force, speed and precision are domains of expertise machines claim that humans cannot.
Let’s Put Some Pieces Together
Yes, some jobs have moved from here in the US to other countries, and not just manufacturing jobs. Yet more jobs just disappeared because automation made them, as the English would say, redundant. More will yet fall to automation, and the numbers will likely astound and shock us. So will coal and steel and other manufacturing jobs stay here, or as promised, come back? Not something you should wager heavily on. Will there be token boomerang jobs that come back, like the Carrier plant last month? Oh for sure, politicians and especially Donald Trump need the theatrics. Those few hundred jobs cost Ohio tax payers $7 million over a decade to come, but will the jobs last that long? Did United Technologies, the parent of Carrier, figure keeping the jobs was a small cost against the loss of its billions in defense contracts? I would wager on that. During much of the Obama administration, and for about the past 75 straight months, private sector growth was about 180,000 to over 200,000 new jobs PER MONTH. Do the math when comparing.
But what about job loss? If government spending on social programs, research, and health care is drastically cut, how will job growth match those numbers, much less exceed them? How will curtailing trade agreements grow jobs instead of eliminating enormous markets for our goods and services that rely on open trade?
It is easy for me to believe that there are very wealthy individuals who will profit handsomely in their business endeavors if deregulation opens up natural resources and gives them a competitive advantage. They may even create a few thousand jobs by doing so. But higher prices will hurt a consumer base whose spending is precariously close to its earnings and their social nets fray and disintegrate with conservative, you’re-on-your-on policies. Go back to oil for a moment. Deregulation might mean more fracking state production and even open up offshore and – heaven help us – Alaska permafrost drilling and production. In a world market with a glut of oil and lower prices than many of us have ever known, crude prices should stay the same or resume their decline. That makes coal even more relatively expensive and makes exploration, drilling and production investments more risky and eventually unlikely.
Will some manufacturing jobs return here? I believe so. They will follow models, like Foxconn and others, and come back in highly automated factories that employ mostly highly skilled and trained information workers to plan and control operations. After all, if you don’t have to pay salaries, it doesn’t matter that wages might be high.
I avoid any mention of climate change/global warming with its well recognized consequences of rising sea levels, heavy rains and storms, and increasing drought in already arid regions. It would fall on deaf ears to many I would like to seriously consider job growth and protection in the next few years, so let’s just say it’s important, in all of these contexts, to some of us.
Said in just a few words, so many of Trump’s potential policies will conflict with each other in terms of his campaign promises to the many people who have lost their jobs, and feel as though the country is slipping away from them. I fear they will be left holding the bag just as empty as the campaign promises made to them.