Rise of the second machine age
Whether or not technological innovation is good for our working lives has been debated since Plato condemned the invention of writing for ruining human memory. Unlike Plato, however, we’re not dealing with tools like pen and paper. Now we’re dealing with machines. In the past, our ancestors lived through the Stone Age, the Bronze Age, and the Iron Age, but today we’re in the Machine Age. More specifically, we’re in the Second Machine Age.
The first and second machine age
The Industrial Revolution was the “first machine age.” It saw the rise of steam locomotives, internal combustion engines, factories, etc. Those technologies did physical work. According to economists Erik Brynjolfsson and Andrew McAfee, we’ve entered “the second machine age.” This age is characterized by computers, automated machines, and artificial intelligence. These technologies do mental work. To quote Brynjolfsson and McAfee,
Now comes the second machine age. Computers and other digital advances are doing for mental power—the ability to use our brains to understand and shape our environments—what the steam engine and its descendants did for muscle power (p 7-8).
That’s from their book, The Second Machine Age. It argues today’s economy is being transformed by three features:
- Exponential improvement of computers: The computing power of circuits has roughly doubled every year since the 1960s (an observation originally made by engineer Gordon Moore—hence called Moore’s Law).
- Digitization of almost everything: Each day, more analog information (books, movies, news, etc.) gets digitized, or turned into computer data.
- Innovation: Organizations use this digitized information to outsource creative work to crowds (“crowdsourcing”). For example, NASA posts scientific problems online and offers rewards to anybody who solves them.
Whether you’re an economist or not, this transformation may affect your livelihood, and Brynjolfsson and McAfee’s book has good, bad, and ugly news about the new economy. Here’s a quick synopsis of The Second Machine Age.
Good news: more bounty
First, the good news. Technological innovation will create more goods and services at a lower cost to consumers (what Brynjolfsson and McAfee call “bounty”). From machine intelligence (smart devices) to e-commerce (online shopping), innovation creates more goods and services at lower prices.
Bad news: fewer winners
Next, the bad news. Unfortunately, most financial gains from this bounty will flow to a tiny group of winners. To illustrate with a simple example, compare construction workers to software programmers. Each construction worker makes money based on, say, how many bricks each lays down. Whoever lays down more bricks makes more money, but all workers will receive paychecks as long as they lay down bricks.
In contrast, not all software programmers will get paid for their work. Whoever programs the best software will end up with all the profits, because nobody wants to pay for the second, third, or forth best software. The result is a winner-take-all market, with few winners and many losers. Which brings us to the ugly news.
Ugly news: technological unemployment
What happens if there’s more bounty but fewer winners? What happens to the losers? Unfortunately, they could remain unemployed. To understand why, consider a past-present comparison.
In the past, when technological innovation destroyed old jobs, it also created new jobs. For instance, the printing press put scribes out of work, but it created jobs like printing-press operators. Therefore, a solution to unemployment is to retrain displaced workers for new jobs.
The present situation, nevertheless, is different for one reason: automation. Automated machines and computers are replacing human workers. In this case, technological innovation may end up destroying more jobs than it creates.
Will machines take my job in the second machine age?
A question we’re all wondering, then, is, which jobs could automation destroy? Although Brynjolfsson and McAfee refrain from forecasting exact unemployment figures, they make a general prediction. Routine jobs are in especial danger. In other words, only non-routine jobs may survive automation.
This routine versus non-routine distinction matters, in particular for anyone who has asked: “Will machines take my job?”
Routine vs. non-routine jobs in the second machine age
Routine jobs refer to routine manual work (industrial jobs such as machine operator, factory worker, and driver) and routine cognitive work (office jobs such as data entry clerk, cashier, and secretary). For better or worse, machines are going to take over many of these jobs.
Non-routine jobs refer to non-routine manual work (service occupations such as hairdresser, janitor, and nurse) and non-routine cognitive work (professional careers such as financial analyst, AI scientist, and teacher or trainer). These jobs are likely to remain in human hands.
For those questioning how automation will affect the future of work (including students contemplating what sort of jobs to pursue), The Second Machine Age should pique your curiosity. People interested in politics or economics also may find the book valuable, especially its tax and policy proposals to mitigate the risk of technological unemployment. For example, the authors propose a negative income tax and increased infrastructure investment.
Whether or not the politicians listen is another story, but they’ll probably be informed if the voters are. In that spirit, feel free to leave a comment below if you have a thought to add, or check out other Recommended Reading on this site.
References
Brynjolfsson, Erik and McAfee, Andrew (2014). The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. New York: W. W. Norton & Company.
This is a very timely post! As a professional in accounting field, I always look for ways to diversify my skill set to mitigate the effects of the inevitable automation. And yet, I still encounter colleagues in my field who are skeptical about it.
I agree with Brynjolfsson and McAfee. Though I wonder, with more goods and services available, but with fewer “winners”, how many people actually will be able to afford these products?
Hi Stan.
You ask a crucial question about what to do when much of the population won’t be able to work or earn enough after the automation revolution (assuming technological unemployment is a risk). There’s a problem if we can’t create enough new jobs to accommodate both a rising population and a shrinking pool of available work. And, of course, if people can’t work, they can’t buy or consume. Honestly, I’m not sure what the solution is, as I’m no economist.
Brynjolfsson and McAfeee suggest a Negative Income Tax (NIT), in which the government would credit or refund taxpayers earning below a certain income level. In effect, NIT would guarantee a minimal income to those not earning enough. Nobel laureate Milton Friedman was a fan of this idea, and I personally find it compelling.
Other have suggested a Universal Basic Income (UBI), in which everyone receives an unconditional stipend from the government. As I understand it, UBI also guarantees a minimal income but without work stipulations. I believe Finland experimented with this idea, but there’s dispute as to how fair or effective it is.
Either way, we need to put pressure on leaders to push for big ideas … and quickly!