Monday, December 29, 2014

Year-End Automation Roundup

Since this has been a fixture on this site over the years, I think it's appropriate to end this year with a roundup of automation/technology stories that have come out over the past few months.

Reading these is proof positive of the old adage that you could lay all the economists in the world end-to-end and still not reach a conclusion.
Clearly, many workers feel threatened by technology. In a recent New York Times/CBS News/Kaiser Family Foundation poll of Americans between the ages of 25 and 54 who were not working, 37 percent of those who said they wanted a job said technology was a reason they did not have one. Even more — 46 percent — cited “lack of education or skills necessary for the jobs available.” 
Self-driving vehicles are an example of the crosscurrents. They could put truck and taxi drivers out of work — or they could enable drivers to be more productive during the time they used to spend driving, which could earn them more money. But for the happier outcome to happen, the drivers would need the skills to do new types of jobs. 
The challenge is evident for white-collar jobs, too. Ad sales agents and pilots are two jobs that the Bureau of Labor Statistics projects will decline in number over the next decade. Flying a plane is largely automated today and will become more so. And at Google, the biggest seller of online ads, software does much of the selling and placing of search ads, meaning there is much less need for salespeople. 
There are certain human skills machines will probably never replicate, like common sense, adaptability and creativity, said David Autor, an economist at M.I.T. Even jobs that become automated often require human involvement, like doctors on standby to assist the automated anesthesiologist, called Sedasys. 
Elsewhere, though, machines are replacing certain jobs. Telemarketers are among those most at risk, according to a recent study by Oxford University professors. They identified recreational therapists as the least endangered — and yet that judgment may prove premature. Already, Microsoft’s Kinect can recognize a person’s movements and correct them while doing exercise or physical therapy. 
Other fields could follow. The inventors of facial recognition software from a University of California, San Diego lab say it can estimate pain levels from children’s expressions and screen people for depression. Machines are even learning to taste: The Thai government in September introduced a robot that determines whether Thai food tastes sufficiently authentic or whether it needs another squirt of fish sauce. 
Watson, the computer system built by IBM that beat humans at Jeopardy in 2011, has since learned to do other human tasks. This year, it began advising military veterans on complex life decisions like where to live and which insurance to buy. Watson culls through documents for scientists and lawyers and creates new recipes for chefs. Now IBM is trying to teach Watson emotional intelligence. 
IBM, like many tech companies, says Watson is assisting people, not replacing them, and enabling them to be more productive in new types of jobs. It will be years before we know what happens to the counselors, salespeople, chefs, paralegals and researchers whose jobs Watson is learning to do. 
Whether experts lean toward the more pessimistic view of new technology or the most optimistic one, many agree that the uncertainty is vast. Not even the people who spend their days making and studying new technology say they understand the economic and societal effects of the new digital revolution. 
When the University of Chicago asked a panel of leading economists about automation, 76 percent agreed that it had not historically decreased employment. But when asked about the more recent past, they were less sanguine. About 33 percent said technology was a central reason that median wages had been stagnant over the past decade, 20 percent said it was not and 29 percent were unsure. 
Perhaps the most worrisome development is how poorly the job market is already functioning for many workers. More than 16 percent of men between the ages of 25 and 54 are not working, up from 5 percent in the late 1960s; 30 percent of women in this age group are not working, up from 25 percent in the late 1990s. For those who are working, wage growth has been weak, while corporate profits have surged.
As Robots Grow Smarter, American Workers Struggle to Keep Up (New York Times)
You've probably read some widespread sillinesses about how technology is moving us toward a world split between "high-skill" and "low-skill" jobs. Worriers claim that people with high-skill jobs will gobble up all of the economic pie, and those with low-skill jobs will be left with mere crumbs. This notion was perhaps best exemplified by economist Tyler Cowen's book Average is Over. 
This is nonsense. Because high-skill jobs are in peril, too. And sometimes, their death will make way for a raft of new "low-skill" jobs. 
For example, look at the future of the general practitioner of medicine. This is considered the epitome of the high-skilled, secure, remunerative job. Four years of college! Four years of medical school! Internship! Residency! Government-protected cartel membership! 
And yet, this profession is going the way of the dodo bird. 
To understand why, the first thing you need to understand is that multiple studies have shown that software is better able to diagnose illnesses, with fewer misdiagnoses. Health wonks love this trend, known as evidence-based diagnosis, and medical doctors loathe it, because who cares about saving lives when you can avoid the humiliation of having a computer tell you what to do. 
Then you need to look at companies like Theranos, which allow you to get a blood test cheaply and easily at Walgreens, and get more information about your health than you'd get in a typical doctor's visit. 
Then look at a company like Sherpaa, whose mobile app provides you diagnoses, helps you get your prescriptions filled, refers you to specialists, and so on. Right now, Sherpaa works with doctors. But there's no reason to think it couldn't eventually work with software (and in the meantime, work with cheaper Indian doctors rather than morbidly expensive American doctors). 
But, you say, we won't be able to get rid of the human general practitioner absolutely. People will still need human judgment, and the human touch. 
You are right — absolutely right. But the human we need is someone with training closer to a nurse's than a doctor's, and augmented by the right software, would be both cheaper and more effective than a doctor.
How computers will replace your doctor (The Week)
In my view, from the economic perspective, the technological forces driving this revolution tend to have the following three downside biases. That is, advances in technology tend to be:
  •     capital intensive (favors those who already have money and other resources);
  •     skills biased (favors those who already have a high level of technical skill); and
  •     labor saving (reduces the total number of jobs in the economy).
The risk is that workers in high-skilled, blue-collar manufacturing jobs will be displaced by machines before the dust settles at the end of the Third Industrial Revolution. We may be heading toward a future where factories consist of one highly skilled engineer running hundreds of machines—with one worker left sweeping the floor. 
In fact, the person who sweeps the floor may soon lose that job to a faster, better, cheaper, industrial strength Roomba Robot! 
For the last 30 years, emerging-market economies have increasingly displaced developed-market economies in the manufacturing sector as a base of production. This is a story we all know: the transition from the old industrial powers of Western Europe and North America to the new ones in Asia. But despite this shift, developed-market economies have somehow made up for those losses in their labor markets. 
Over the last 20 years, the overall unemployment rate in the United States has hovered around 5% on average—except during periods of economic recession, when it has spiked upward for short periods of time. 
In general, however, the loss of those manufacturing jobs has not caused catastrophic levels of unemployment. 
How? Well, the short answer is the service economy. 
(Of course, this replacement of manufacturing jobs with service jobs has not been equally distributed. Some regions have suffered more than others. For example, the so-called Rust Belt in the upper Midwestern section of the United States has experienced more economic pain than most other regions. But while the local suffering has been great in those regions hardest hit, the overall trend throughout most developed-market economies is that lost manufacturing jobs have been absorbed largely by new jobs created in the service sector.) 
In my view, however, there’s no guarantee that this positive scenario—of service-sector jobs making up for lost manufacturing sector jobs—will continue.
Rise of the Machines: Downfall of the Economy? (Nouriel Roubini)
The signs of the gap—really, a chasm—between the poor and the super-rich are hard to miss in Silicon Valley. On a bustling morning in downtown Palo Alto, the center of today’s technology boom, apparently homeless people and their meager belongings occupy almost every available public bench. Twenty minutes away in San Jose, the largest city in the Valley, a camp of homeless people known as the Jungle—reputed to be the largest in the country—has taken root along a creek within walking distance of Adobe’s headquarters and the gleaming, ultramodern city hall. 
The homeless are the most visible signs of poverty in the region. But the numbers back up first impressions. Median income in Silicon Valley reached $94,000 in 2013, far above the national median of around $53,000. Yet an estimated 31 percent of jobs pay $16 per hour or less, below what is needed to support a family in an area with notoriously expensive housing. The poverty rate in Santa Clara County, the heart of Silicon Valley, is around 19 percent, according to calculations that factor in the high cost of living. 
Even some of the area’s biggest technology boosters are appalled. “You have people begging in the street on University Avenue [Palo Alto’s main street],” says Vivek Wadhwa, a fellow at Stanford University’s Rock Center for Corporate Governance and at Singularity University, an education corporation in Moffett Field with ties to the elites in Silicon Valley. “It’s like what you see in India,” adds Wadhwa, who was born in Delhi. “Silicon Valley is a look at the future we’re creating, and it’s really disturbing.” Many of those made rich by the recent technology boom, he adds, don’t seem to care about “the mess they’re creating.” 
The wealth generated in Silicon Valley is “as prodigious as it has ever been,” says Russell Hancock, president of Joint Venture Silicon Valley, a nonprofit group that promotes regional development. “But when we used to have booms in the tech sector, it would lift all boats. That’s not how it works anymore. And suddenly you’re seeing a backlash and people are upset.” Indeed, people are stoning buses transporting Google employees to work from their homes in San Francisco. 
The anger in Northern California and elsewhere in the United States springs from an increasingly obvious reality: the rich are getting richer while many other people are struggling. It’s hard not to wonder whether Silicon Valley, rather than just exemplifying this growing inequality, is actually contributing to it, by producing digital technologies that eliminate the need for many middle-class jobs. Here, technology is arguably evolving faster than anywhere else in the world. Does the region really portend a future, as Wadhwa would have it, in which a few very rich people leave the rest of us hopelessly behind?
Technology and Inequality The disparity between the rich and everyone else is larger than ever in the United States and increasing in much of Europe. Why? (MIT Technology Review)
 ...But a new study by professional services firm Deloitte has quantified the rate of destruction for the U.K. jobs market over the next 20 years – predicting that around one-third (35 per cent) of existing jobs across the U.K. are under high risk of replacement via automation over this time period. 
The study authors claim those the lower paying jobs will get laid off at 5 times the rate of higher paying jobs. During the industrial revolution demand for factory workers rose as machines became more productive because the value of human labor was enhanced by what the humans, using machines, could produce. But in the current era the machines are getting smarter and more autonomous. 
What is telling: chart 4 shows that since year 2000 in manufacturing employment has increased for advanced degree holders while going down for everyone else. The size of the decline is most severe for high school drop-outs. This pattern is going to repeat in a growing list of industries. 
Some dark factories (a.k.a. lights-out factories) have no humans on the factory floor for substantial lengths of time. Technicians still come in to do repairs or rearrange the equipment. The rest of the time machines do all the work. Robots are moving into more industries. Dark warehouses are coming next. Distribution is getting automated in warehouses. 
    Such advances in manufacturing are also beginning to transform other sectors that employ millions of workers around the world. One is distribution, where robots that zoom at the speed of the world’s fastest sprinters can store, retrieve and pack goods for shipment far more efficiently than people. Robots could soon replace workers at companies like C & S Wholesale Grocers, the nation’s largest grocery distributor, which has already deployed robot technology. 
We can not simply extrapolate from a graph of the last N years to say what the world will be like 20 years from now. Discontinuities happen. Suddenly some development will depart from trend. This could mean that robots do not advance as some expect. But with many new kinds of gadgets there is an S-shaped curve where suddenly adoption goes up sharply when a technology reaches a level of maturity that makes it broadly useful. This has happened with PCs and cell phones for example. 
My expectation: The ranks of unemployed and not looking for work will continue to grow. The work ethic is weakening and demand for government support to enable a non-working life will grow as the work ethnic declines.
In UK 35% Of Existing Jobs To Be Automated In 20 Years (Future Pundit)
The U.S. unemployment rate had declined back to the range of 5.0% by August 1964, but concerns over how the U.S economy might adapt to technology and automation remained serious enough that President Lyndon Johnson signed into law a National Commission on Technology, Automation, and Economic Progress. The Commission eventually released its report in February 1966. when the unemployment rate had fallen to 3.8%.
Before reviewing the tone and findings of the Commission, I'll just note that when I run into people who are concerned that technology is about to decimate U.S. jobs, I sometimes bring up the 1964 report. The usual response is to dismiss the 1964 experience very quickly, on the grounds that the current combination of information and communications technology, along with advanced in robotics, represent a totally different situation than in 1964. It's of course true that modern technologies differ from those of a half-century ago, but that isn't the issue. The issue is how an economy and a workforce makes a transition when new technologies arrive. It is a fact that technological shocks have been happening for decades, and that the U.S. economy has been adapting to them. The adaptations have not involved a steadily rising upward trend of unemployment over the decades, but they have involved the dislocations of industries falling and rising in different locations, and a continual pressure for workers to have higher skill levels.
Automation and Job Loss: The Fears of 1964 (The Conversible Economist)
Skill unemployment is on the rise in the US and the UK, where it doubled between 2000 and 2012, and in Spain and Italy, where it tripled in the same period. In Germany and France we do not see a rise in skill unemployment. The doubling of academic unemployment between 2000 and 2012 and the modest increase in the skill premium in the US, in spite of the slow expansion of education, suggest that capital bias technology is driving this outcome. In Germany, skill unemployment is low and did not increase between 2000 and 2012 precisely because education was advancing slowly there. 
Are we fighting the wrong battle? Is the rapid expansion of education as an answer to the challenges of globalisation the wrong way to go? Will we see an excess supply of human capital in the next decade that earns penny wages? It may well be that the ‘war for talent’ and the scarcity of human capital is an issue of the past.
Globalisation and the rise of the robots (VoxEU)
A few weeks ago McDonald's announced new technology “to make it easier for customers to order and pay for food digitally and to give people the ability to customize their orders”. Think mobile ordering and self-ordering kiosks. Chili’s Bar & Grill is starting a tablet ordering system and Applebee’s is doing the same in 2,000 of its restaurants. In China a fully robot-staffed restaurant (yes, they cook too) has been running for a few years. There’ll be no real people left behind any fast food counters in 15 years.

This is partly about cutting the cost of labour. The clamour for higher minimum wages is rising (it has just gone to $15 an hour in San Francisco) and savvy employers know that even without legislation, wages won’t stay static in the west for ever. Wage growth might be sluggish here, but in the US worker compensation was up by an annualised 2.2 per cent in the third quarter. 
It’s not just about nasty capitalists trying to avoid paying a living wage: ordering via machine is quicker and better. Robots don’t get sick, get pregnant, ask for more money, argue about repetitive tasks, or fall in love with their co-workers. They might also be better sales people than people. Are you more likely to sign up to a loyalty programme on an iPad in your hand while you wait for an order, or on a form shoved at you by a harassed teenager at the counter? 
Nor is it just about restaurants. According to research company Gartner, one in three jobs will be replaced by some kind of software, robot or smart machine by 2025. Visit a factory today and you can see how. I went to a sawmill last year. From the second the logs rolled off the lorry the only human interaction came via a computer control room. They were sized, sorted and cut on orders from a software system. Amazing. 
Robots are already cleaning hospitals, vacuuming the houses of people slightly richer than you; picking strawberries in Japan (with a camera that analyses ripeness). But this also isn’t just about boring jobs. Take what you can when you can, said one strategist to me a few days ago when I was wondering whether to write more or not. “Give it a decade and both our jobs will be being done by a machine.” Far fetched? Not really. 
Futurist Ray Kurzwell reckons that by 2020, computer processing power will have reached the level of a human brain. Middle-aged bankers might want to look back to their first graduate job: how much of it is automated now? A third? We don’t need junior number crunchers any more. 
It’s the same in journalism. I had two assistants when I first started out as an editor. Now I have an iPad. I miss the girls. But there it is. The same goes for train drivers: I was mildly surprised to see last month that the trains at the Tokyo Disney Resort have no drivers, but run very smoothly. We also don’t need so many cameramen or pilots – aerial footage of everything from war zones to farmland comes from drones. And what of nurses? If remote devices can monitor patients with heart disease and diabetes and infrared light can guide robots to the right veins to take blood from, we don’t need so many of them either. The same goes for surgeons.
Making Money in an age of Machines (Financial Times)
There are a few more reasons to be skeptical about secular stagnation. For one, productivity growth might not be as bad as we think. That's because the way we measure it is having a harder and harder time capturing the real gains in the economy. Just think, for example, about the way smart phones are replacing books, newspapers, cameras, scanners, bank ATMs, voice recorders, radios, encyclopedias, GPS systems, maps, and dictionaries, among other things. All of these items used to be sold separately and each counted as a part of GDP. Now, many of them are available as free apps, and aren't counted as a part of GDP. This under-measuring of economic activity translates into slower observed productivity growth. And this problem is only going to get worse as more of the economy becomes digitized. 
Actually, a tech slowdown could be the least of our problems. As Erik Brynjolfsson and Andrew McAfee argue in The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies, it looks like smarter and smarter machines are going to radically transform the economy. This digitization of everything could mean a big growth is just around the corner. The near future will be an economy where machines like IBM's Watson will be coordinating driverless cars, diagnosing illnesses, and organizing robot work in homes. So productivity growth and returns on capital will both be high. The real challenge, according to Brynjfolsson and McAfee, won't be too little economic growth, but dealing with the disruptive nature of so much of it. 
And finally, if demographics really are destiny, then our destiny isn't secular stagnation. As Bill McBride has pointed out recently, the Census Bureau now reports that Baby Boomers aren't the largest cohort anymore, and that the prime working-age force is expected to start growing again in a few years. In other words, the demographic outlook actually points to stronger economic growth. (Thanks, millennials). And this isn't just an American development: the United Nations projects that working-age populations will increase in many parts of the world through 2050. That should create plenty of demand for new investment. 
In short, it doesn't seem like secular stagnation is the right story for the U.S. economy. A better story is that the economy got hit by a once-in-three-generations crisis that's taken awhile—too long, really—to overcome. But in the long run, the slump will be dead. And it might just be "Morning in America" again.
Here’s why Larry Summers is wrong about secular stagnation (Washington Post) Wow. Talk about delusional. You'll notice he is celebrating the increasing number of people entering the workforce immediately after a paragraph about how jobs are being eliminated through technology. Yes, ladies and gentlemen, this is how people involved in the "science" of economics think. And these people are writing op-eds in our media! In the comments section several people point out this intellectual absurdity.


  1. I once lost my job to automation, so i moved to a beach in mexico and surfed for a few years. Best experience of my life.

    How are you feeling these days?

    1. Nice. How did you pull that off? I wouldn't mind taking a few years off, but I don't know what I'd do after.

      Felling pretty good, thanks. The Bell's Palsy is gone as far as I can tell..

    2. Glad you're doing better.

      At the time I was single so I sold most of my possesions, and bought an old van and drove around the pacific coast until i found the perfect beach. I ended up renting a palapa and just living day by day.

      It did take some adjustment getting back into the American culture and finding a job again, but what the heck, life is a journey...

  2. Christmas day, some of the millennials in my family were bragging up automation until I pointed out to them that they were the ones being automated. I used the supermarket self checkout as an everyday example. You have to scan the items, punch in codes, do your own bagging and half the time a glitch pops up and you need to stand around helplessly waiting for an employee to rescue you with her secret codes. If you go through the regular checkout a nice person does all that for you and you get a free smile at the end.

    1. What's weird is how people, especially Americans, assume all innovation is going to benefit them personally, no matter who they are, no matter the evidence. And they also seem to be unable to see the downsides. Partially that's because of the media which is there to encourage people to believe in the latest corporate marketing schemes. Partially it's because of all the techno-miracles that Americans especially have seen in the last few generations. But this time it may be different.

      Even scarier, Americans' hidebound Calvinist attitudes to work and laziness are going to be toxic in this new automated environment.


Note: Only a member of this blog may post a comment.