Martin Ford, a “Silicon Valley entrepreneur” and the usual sort of progressive grumbler that complains about the capitalism that made him wealthy, has been on a tear in the news and with his latest book Rise of Robots: Technology and the Threat of a Jobless Future about the rise of automation in the workplace.
It’s been about 600 years since people started having the vapors over automation taking their jobs — from the saboteurs of 15th Century Holland, to the Luddites of the early 18th Century, to the labor unions and the science fiction dystopias of 20th Century, we’ve been told that automation — the use of labor saving devices, would steal jobs and impoverish, well, everyone. Panic! Run for your jobs!
Except that’s not what’s happened.
At every turn, the addition of automation has changed the work force, not constricted it. The creation of mechanical labor-saving devices creates whole systems around their construction, sale, usage, and maintenance, all while freeing people from (at least one form of) drudgery, enriching laborers in general, and freeing them to focus on other endeavors.
Ford, however, assures us that this time the end is well and truly neigh! His reasons:
1) Super-smart Oxford academics inform us that “47%” of US jobs are in danger of automation! Half of us will be out of work!
Problem 1: The work of Carl Frey and Michael Osborne is, by their own admission, subjective. It’s a thought piece with no justification or proof for its assertions, as doomsaying predictions are 90% of the time. (A shocking percentage I wouldn’t have believed, if I hadn’t just made it up…kind of like Frey and Osborne.)
2) Ford sees stagnant wages and rising inequity of income and promptly blames this on evil capitalism.
Problem 2: What he doesn’t see are several trends that explain portions of this — like the collapse of the dollar, certification inflation (where the degree you had 30 years ago is now worthless because a large cohort of college students from the last 20 years got the same piece of paper), a shift in the type of labor — from industry to information, and the creation of a global market where competition against cheaper labor pushes labor prices down in some industries (like his own tech sector.)
3) Computers are getting cheaper, more powerful, and smarter AI is helping them to do things that only science fiction writers would buy into last century. Google language translation apps and self-driving cars particularly forge the fear of Skynet in him.
Problem 3: They are…and the sorts of things they can do well will certainly improve, but lets address the two examples above first: Translation programs, right now, are pretty hit or miss, but improving. This is not going to put humans out of jobs in most places where translation requires the understanding of non-verbal queues, of slang, etc. Could they eventually reach the point where we have language-saavy androids officiously doing translation for useless UN politicians? Sure…but it’s unlikely. As for self-driving cars, there will be a big rush — I suspect — to roll out Google Taxi or whatever until the first really disastrous accident. Because machines fail sometimes. But say they don’t — how much more work could you get done, or how much more meaningful time with your family could you have during a commute if you didn’t have to worry about the texting old person (and yes, they are the largest demographic to text and wreck…look it up or have your Google data miner do it) going 10 miles under the speed limit..?
Ultimately, Ford’s screed is less about robotics and automation than it is about how awful income inequality is, and how evil capitalism and their robot minions will create a perpetual underclass of people that can’t get work. It’s a great premise for a science-fiction movie…but we’ve seen that movie before. Here’s how creative destruction really works…
Example the first: The creation of the cotton gin made it much easier to pull cotton from its husk than prior to the machine’s invention. It allowed textile manufacturers to make more cotton for more and cheaper clothing. It destroyed a particular job — separating cotton, but it made for other labor…namely picking cotton, and most likely led to the survival of slavery in the United States into the mid-1800s. But it also meant more people — mostly women, who could not find employment that didn’t involve lying on their backs or cleaning and cooking — to work on clothing manufacture.
But Scott! Those women were exploited! Here’s a good question…if they had it so damned good on the farm, darning clothes, cooking, pumping out and raising kids, all while working in the fields with their husband and worrying that the crop might not be enough to survive or pay the landlord…why go to work in the “dark, satanic mills” of the city?
It also created jobs to build the gins. To market, sell, and transport the machines. To train people on them. To fix them. And those jobs were in demand, required specialized training, and paid better than the previous job hulling cotton.
Example the second: The automobile, from the start, had aspects of automation — namely the assembly line. Portions of the automobile were rolled from one station to the next to be constructed by laborers. They made good money…mostly so they could afford the product they were constructing. The car created jobs in mining, in metalworking, in mechanics, in marketing; it created entrepreneurial opportunities in selling cars, in fixing them. It created jobs for the roads that needed to be improved, for those people to pump gas into them (not to mention the just extracting oil and refining it.) It created licensing schemes and insurance schemes — all requiring people to work. Folks washed the cars. Created upholstery and aftermarket parts to improve performance, which required people to work in engineering fields….and most of these jobs, gave low skilled workers entrance into labor (the car washer or road navvie, say), or gave those who trained in this in-demand skills jobs that nearly always paid better than those they had previously had (engineerings, car construction hands.)
As the automobile manufacturing matured int he 1970s and ’80s, robots joined the assembly line, and destroyed certain jobs…but created others. Maybe you didn’t bolt the bumper on, but a worker checked the quality. Less manual labor, but you still got paid well. Others had to calibrate, fix, and program the machines. Those machines created jobs on other assembly lines where they built the machines.
The creative destruction of automation might “take jobs” from a select group, but it creates other opportunities, and often those jobs were more skilled, more specialized, and hence better paying than the ones before. If you lost our job and didn’t retrain for another, this is not a reflection on capitalism, or evil entrepreneurs (like Ford), but on the worker who cannot adapt to his surroundings.
Ford’s assertion that income inequality is the main ill of capitalism shows the usual lack of understanding of historical trends across time. Since the industrial revolution began, income throughout the developed world (and we have an excellent live-time case study in China’s labor force) rises across all quintiles of income. The rich get richer, and usually at a higher rate, but they are also the ones that rise capital and create the opportunities than employ others…rising their level of income and prosperity.
Let’s just look at the period that Ford is concerned with:
The only quintile not seeing a rise is…none of them. The lowest quintile is pretty static since 1995, but if you are locked into an income of less than $20,000 a year, that’s not evil capitalist conspiracies doing it…you are making real bad decisions regarding your employment. Hell, if you work hard at McDonalds it’s really hard not to get promoted. All other quintiles showed steady improvement. the reason it doesn’t feel like it is that the government has been destroying the value of your money since we went off the gold standard in the 1970s (thanks, Nixon!)
That gives you this…
Note the drop in real income is not the poor…it is in the middle class — that nemesis on most political ideologies thanks to their having done a real day’s work and who have enough money to feel that the politicians should listen to them. (And they should.) Even the “rich” are taking a hit here.
It is the lack of understanding of the most basic of modern economic influences, and an ignorance of historical trends, that is why Ford’s evidence is so shoddy, and his premise is laughable. (Until the Terminators kill us all…)
Funnily, Ford misses the most likely big technology trend that will impact manufacturing — 3D printing:
The rise of the do-it-yourself manufacturing will hit those industries that can’t or won’t the demands of modern consumers for configurable,individualistic products. Even this technology will require people who can design and build 3D printers, who can program them and fix them. It will need people who conceive of products for people who do not want to take the time to build them for themselves…the same division of labor that has been driving economies since the days of the Sumerian city-states.