Baby Boomers and robots
There has been much discussion in recent years concerning the aging of the enormous “baby boom” generation (birthdates 1943-1960 in the U.S.) and the crisis it will spawn in the economy. According to a recent article published by the Levy Institute
written by Walter Cadette, during the 1970s the ratio of workers to retirees was four to one. Today it is three to one, and it will be two to one by 2030. In other words, for every two working people there will be one retiree who will require economic and personal support.
The situation is not so bad in the U.S. as it might have been. Between 1984 and 2000 America experienced a second baby boom, originally referred to as the “echo boom” but now called the “Millennial” generation. During this boom the number of children per family rose significantly from its all-time low in the mid 1970s – in other words, U.S. Boomers wanted more children than the previous “Silent Generation” during the 1960s and 1970s. U.S. birthrates popped several times over the “replacement” level of 2.1 children per family in the 1990s. This is little more than half the family size during the peak of the original 1950s boom, but quite significant.
In contrast, birthrates in Europe and Japan stayed low and then fell. In most countries in Western Europe, birthrates are currently running around 0.8 children per family – only 1/3 the level needed to replace the current population. There is no sign of an upturn – indeed, birthrates continue to fall. Ultimately, this will cause an overall population decline, but long before this happens, these countries will be faced with an impossible situation of one worker (or less) per retiree.
Think of it: circa 2020 every working person in Western Europe and Japan will be totally subsidizing a retiree. Consider that these countries have more extensive socialized healthcare. They are up against an impossible wall.
The Levy article demonstrates that things are even worse than this. Most discussions of the effect of aging Boomers have centered on Social Security and Medicare. The implicit image is a “young at heart” Boomer who is retired but is still playing golf, buying sports cars, and getting stoned at “geezer rock” concerts. This may describe the world of 2015, but not 2030. By that time, the huge Boomer population will be physically frail and need comprehensive support from caretakers, prescription drug manufacturers and hospital workers. There will be an enormous need for long-term care, typically very expensive because multiple caretakers are needed. Instead of a Boomer popping their meds while swinging their tennis racket, they will sit quitely as nurses, physical therapists, doctors, pharmacists, etc. tend to their ever-expanding needs.
Economically, this is an unworkable situation. At-home care for the elderly runs in excess of $100,000 in today’s U.S. dollars. Consider that most elderly die after several years of intensive support. How can the bill be paid for if only 1-2 people are working per every retiree? According to some studies, taxation would have to double to levels in the 70% range in the U.S. and 100% in Europe – totally impoverishing the younger workers at the expense of the old.
Even leaving aside the economic issues, this is an equally impossible situation in terms of personnel. Quoting Cadette”
“How does a nurse’s aide bathe, feed, and tidy up (a euphemism if ever there was one) nine or 10 residents with advanced dementia-and hug them too? How do even heroes and saints do a job so demanding? ”
With only one working person per retiree in Europe and Japan, will that person work as a nurse for the elderly? Will the entire working population of Europe and Japan consist of elderly caretakers? Who will run the rest of the economy? We are up against wall.
In this light, the claim that Honda and other robot manufacturers have been making – that their robots are intended for elderly care – does not seem flippant. In fact, robots offer the only solution to long-term care of Boomers as they near the end of their lives. Provided sufficient energy is available (and that’s not a given), we can crank out armies of mechanical nurses to replace the missing human workers.
As computing power has increased, robots are finally becoming real. Even if we assume that the PC era’s “Moore’s Law” fails (computing power doubling every 18 months) the robot of 2020 will be hundreds of times smarter than those today. This may not be enough for the philosophical angst found in robot movies but it is likely enough for a robot to change the bedding or administer medication.
There is another reason robots that jump will concentrate in healthcare. There will be resistance to robots taking over “everyday” jobs like car repair – this has been one of the long-standing robot fears. But elderly care is another story. These are low-paid, thankless jobs. Nobody will want a robot next to them on the job, but they’ll think it is just fine to have one is wheeling Grandma around the ward.
There is a widespread understanding of this future in the medical community. While the “tech” industry doesn’t get that robots are a monster growth industry, healthcare does. In the U.S., the few practical robots currently being constructed are being put into healthcare environment. Compared to Japanese robots they are primitive. Usually, the consist of a wheeled vehicle with a computer monitor at eye level. This monitor can display a simulated doctor’s face (ag) or a real doctor elsewhere in the world. The robot can access the hospital database and guide people to their morning EKG test.
A good example of a company producing this kind of robot is Intouch, which estimates the elderly population will double in the next 10 years. Today, there are over 400,000 unfilled nursing positions – how bad will it be in a decade? Their “wheels and monitors” robot (called Dr. Robot) is currently being tested at Johns Hopkins. At present it mostly supplies “remote presence” to doctors but this will change quickly. Groups in Japan (the Robita project) and in the U.S. (Sandia) are hard at work developing robots that can handle the simple give and take required of a receptionist. In a decade this could easily evolve into the soothing commentary provided by healthcare givers. Honda is already renting out the Asimo as a primitive receptionist. I predict that in a few years Honda will give a demo of a fully-functioning receptionist capable of much of the work done by healthcare interns (the ones they used to call “candy stripers”).
Contrast this vision of future robots with the one heard in tech media – fantasies of warrior robots. In fact, it will be far easier to create robot nurses than robot soldiers. This is due to the complexity of their respective tasks. A caretaker robot can be “always good” and follow whatever version of Asimov’s robotic laws we want. Its simple goal is to “do no harm.” In contrast, soldier robots must be good and “bad” – and have the wisdom to decide when each behavior is appropriate.
There is a final, ironic note to all this. Watching the HRP robot shuffle around, a student of mine commented that it walks “like an old man.” He’s – the robot has co-ordination comparable to an elderly person. In addition, the motors running its legs can’t develop the torque of human muscles – it is physically weak in a way many elderly people are. Surrounded by their young masters, robots today are like old men in their “second childhood” being helped to walk by their human nursemaids.
In a few decades the situation will be reversed – the robots will turn ever more strong and capable while their charges become weak and frail. The child will lead the parent.