Will grandma want a robot seal?

3
【 This post was originally published on here. 】

Hey Kev. Mom hasn’t opened the fridge or the bathroom door since Tuesday, and she unplugged her diaper sensors. The floorboards aren’t giving off electronic “creaks,” and Jitterbug shows she hasn’t been at church, or Safeway, or the gym. I ought to just call but I’ll admit I’m scared—you know how she is. Last week, I tried to get the concierge to install CareEye cameras in the apartment—you should have heard her, cursing a blue streak. You can take the gal out of Wyoming, but … you know the drill. Same as when we tried to get her into that assisted living joint—everybody we send her is “stupid” (overworked and staring at their screens while they talk to her, probably); the care robots “give her the creeps.” BTW, that squealy seal companion thingy you got her for Christmas? It ran out of juice, or maybe it’s in a closet. Or, I don’t know, maybe she put it on an ice floe and sent it out to sea. Or clubbed it to death. Japanese moms may go for that stuff, but ours won’t…

We’re watching it crest in Japan now, and 30 years from now it will come crashing down upon us: the Silver Tsunami, that vast demographic upsurge that will populate the United States with 100 million people over the age of 65 — twice as many as today.

Life won’t be easy for the millennial generation, sandwiched between the needs of their own children and their aging parents. Getting old will still be difficult, as will watching people get old in a changing country. As the number and percentage of elderly grow, a shortage of caregivers is a near-certainty — especially if we tighten our policies on immigrants, who represent a large proportion of caregivers today.

Medical care — mostly for older people with chronic illnesses — already consumes a sixth of our GDP, and that’s almost sure to rise as the population shifts away from working-age people and toward retirees. The solutions to this problem, if we are to solve it even partially, will have to be numerous, from new preventive-health policies to cost controls on health care. But one avenue that experts and companies are investing great hope in is technology designed around aging.

The human interactions of health care — the time of doctors and nurses and orderlies, the physical challenges of getting someone to the hospital and back, the reaching out between visits to make sure a patient is taking his or her medications and has heat and electricity and food in the refrigerator — are one of the reasons it’s so costly. So far, medicine has been highly resistant to the type of technological efficiencies that have lowered prices in other industries. But that’s starting to change, and one of the key areas is old age.

Experimentation with tech has already begun. Amazon has sold five million of its Echo devices, and it’s marketing them for patients with dementia (unlike humans, Echo never grows frustrated with questions like, “What day is it today?” or “Have I taken my medications?”). In Japan, caregiving robots like Asimo have been employed for several years to talk and sing and lead exercises for residents of nursing homes. Pill packs that beep when you didn’t open them are being marketed now; a few physicians are prescribing blood pressure cuffs and wireless scales that send charts to the doctor each week. Hundreds of millions of Fitbits and Jawbones and Apple Watches monitor heart rates and steps and the weather, and can also be programmed to nudge you to get off your rear end. AARP releases videos aimed at preventing doctor visits by showing caregivers how to change surgical dressings or IV drips. A company called HealthSense provides sensors to assisted living centers that allow them to track whether tenants have been sleeping or eating or going to the bathroom.

The Skype-like practice of telemedicine is also growing in popularity — Teladoc, a Dallas-based company, just announced its two-millionth doctor-patient visit over the phone. The appeal of remote consults is growing as medicine begins to shift from traditional fee-for-service, in which a doctor gets paid for each visit or operation or procedure she does, to “value-based” medicine, where maintaining the patient’s health is both the financial and ethical objective. Kaiser Permanente, the gigantic nonprofit hospital chain that pays its doctors salaries and has led the way to more holistic care, conducted more than 110 million remote patient-doctor interactions last year — via smartphone, videoconferencing and other technology. Slightly more Kaiser patients saw doctors electronically than in person. Scientists are figuring out new ingenious ways to use data to help monitor and care for people. Already, a few sensors on an iPhone — measuring steps, heartbeat, finger-tapping, voice recording — can tell a research scientist whether someone is developing Parkinson’s, whether it’s worsening, what times of day are worst and when the main treatment, L-Dopa, works best. The sensors, monitors and robotics are almost ready for prime time; there’s little technical reason their use couldn’t explode in the near future in health care. “Once something starts to work in technology, it gets better, faster, smarter and cheaper really fast,” says health technologist John Wilbanks, who goes by the lofty title of chief commons officer at Sage Bionetworks.

But there are many institutional obstacles to tech’s use in aging and the health care business—obstacles that didn’t exist for Apple when it transformed the music industry, or even for Tesla and other automakers as they try to introduce driverless cars. Americans don’t pay directly for health care — they get it through insurance. And insurance, whether private or governmental, is shy about paying for new tech, however shiny, that hasn’t been carefully evaluated for safety and value. Eventually, a combination of evidence, political lobbying and cultural shifts can bring about adoption of less-than-perfectly tested technologies, of course. When they do, we will see whether the tech really makes aging better, or just produces different forms of unhappiness.

AS THE DIGITAL health industry grows, one of the standard critiques is that the 20-something whiz kids who design most of the products have no idea what it’s like to be an 80-something trying to use them. Google “aging” and “technology” and you’ll find as many startups aimed at preventing aging as trying to ameliorate it. However, several academic centers — at Johns Hopkins, Stanford, the University of California, MIT and elsewhere — are trying to bring together techies with doctors and their geriatric patients to develop technologies that hold real promise for the elderly. AARP’s policy institute has a project devoted to incorporating tech into aging care.

It isn’t as if technology were an entirely new concept for elder care; the bigger problem has been getting seniors to use it. The LifeAlert pendant (“I’ve fallen and I can’t get up!”) has been around for three decades and the newer versions allow disabled folks to push a button and reach operators who can discuss their problems, recommend an electronic doctor’s visit via telemedicine, or call 911. It’s not complex. But only about a fifth of the owners of the LifeAlert pendants are wearing them during falls, according to robotics expert Tandy Trower. That means that four-fifths of the time, whoever buys them (typically the owner’s daughter or daughter-in-law) is wasting their money.

Even older devices — hearing aids, canes and dentures — go unused all too frequently, notes Bruce Leff, director of the Center for Transformative Geriatric Research at Johns Hopkins University. “Part of it is vanity. Part of it is not wanting to appear impaired, part of it is not wanting, or being able to spend money on these things.”

“A walker is a great bit of technology,” he adds. “That’s another one I can’t get my patients to use.”

Health IT innovators don’t want their gadgets to gather dust on shelves, either. Engineers and data scientists are rapidly introducing new ways to gather terabytes of information on everything from heartbeats to brainwaves to glucose levels. So far, however, the technology enabling us to measure things is still outrunning our ability to make sense of the measurements. Many of the sensor technologies, like the ones that can go in the floor or a shoe, to measure the steadiness of an old person’s gait, rely on computer designs that correlate certain data with, say, deteriorating motor skills. But most of them aren’t yet actionable from a medical perspective. “Clinically, I see people who have unstable gaits and go months without a fall,” said Leff. “Or they look fine, but then they stumble and break their hip.”

This is part of the reason that these gadgets and apps, plentiful as they are, have yet to transform health care and aging the way Amazon transformed shopping, Uber overturned the taxi business, or Apple flattened the music industry. There are other obstacles, too. The FDA regulates some of the technology, and some of it requires the buy-in of doctors, a skeptical lot — the double-blind clinical trial, time-consuming and expensive, remains the gold standard for accepting new devices into medicine. In any case, doctors can’t spend a lot of time with a new gizmo or software unless insurers agree to pay for it and for their time. “There are so many points of resistance,” said cardiologist Eric Topol of San Diego, a prolific author on the coming benefits of health technology. “The incumbents don’t want this to work because it’s shifting responsibility to patients and their caregivers.”

Robotic things have replaced workers in car factories, grocery checkout lines and parking garages, but meeting the needs of an elderly person turns out to be a complicated matter. “It’s easy to watch WestWorld or a Star Wars movie and imagine what things could be like, but when you are actually creating that kind of technology, you learn how hard it is,” says Trower, a former Microsoft senior engineer who’s designing an autonomously mobile, socially interactive companion robot. It’s hard as hell, he says.

“There’s a lot of talk about the rise of artificial intelligence, but I find that it still takes a large measure of human intelligence to design and code this,” Trower says. Teaching a robot language is not just a matter of turning digital words into sounds. There are all kinds of context, facial gestures included. “If I grimace after you tell me a joke, it doesn’t always mean I am angry,” he said. “But at times it could mean I’m depressed. AI takes large datasets to learn and that doesn’t guarantee it learns the right thing.”

Robot designers also wrestle with a design issue known as the “uncanny valley,” which amounts to this: making robots seem human is not necessarily a good idea. “The more human you make the robot look, the greater the expectations that its behavior matches,” said Trower. That’s one reason the developers of robot companions often turn to animals as models, such as the cuddly Paro therapeutic harp seal currently used for dementia patients.

As much as we may shake our fingers at them, people who didn’t grow up with information technology — and even some of us who have — may not welcome its pervasive touch in their lives. Unlike workers at car plants or groceries or parking garages, the elderly don’t have to adopt technology unless they want it. A survey last year showed that only 7 percent of caregivers — professional or familial — were employing health IT to help them with their parents, clients or sick friends. Three-quarters of them said they wanted to use more. But there might be some resistance. “I joked with my parents about [setting up cameras in their apartment],” said Paul Sonnier, a health tech publicist in California. “They said, ‘No way we’ll have that in our house.’ And I said, ‘Well, would you prefer to be in a rest home?’”

The bright side, from Trower’s perspective, is that tech-resistance will be a lot lower when generations raised on the Web and smartphones are elderly.

WHAT WILL THOSE generations encounter as they get older? Just extrapolating from current technology, it’s no stretch to imagine that soon there will be high-tech goggles that allow the home-bound to visit Venice, or a brothel, or Australia’s Great Barrier Reef. An intelligent coffee maker won’t just make your coffee, but can give you a nudge to brush your teeth. A mirror on the wall will scan your face for warning signs of depression, heart attack or stroke. A smart toilet — yeah, yech — will detect abnormal blood levels in your effluvia. Clothes that detect activity levels and heart rates, or cushion your fall, or strengthen your gait are under development. Robots, or some kind of home AI system, will learn to sense your mood and tell your daughter you’re being surly. Self-lacing shoes for the arthritic, glasses that spell out signs for the legally blind — both are on the market — and, a little farther off, robots that can walk you to the bath, put you in it, lift you out and towel you off.

Out in Silicon Valley, up at MIT, and in other places where venture capitalists back engineering they think will reap profits, they are designing robots that can roll through a home to check on your mood, review your vitals and even chat with you before bedtime. Self-driving cars to take you where you need to go. Retirement homes prewired with monitors running in the background, passive technology, invisible to the user, a windowless observatory of your well-being. “You can digitize an elderly person’s mood now,” says Topol.

But how much of this stuff are we going to really want, as individuals and a society? Technological solutions, to be sure, are tempting to the caregiver. “Between a phone and a webcam and thermostat you can measure someone’s environment relatively capably, and if you get older people not living in the same city with their kids you can imagine that being effective. You can imagine me checking in on my dad,” said Wilbanks, whose father suffers from cancer. “It’s a safe assumption that this is going to happen but we don’t know what the contours will be like. Will it be creepy panopticon, or Kumbaya? Unfortunately our culture trends toward panopticon.”

One could imagine, Wilbanks said, “the typical Silicon Valley solution — ‘I’m measuring John, he seems to be developing Parkinson’s, so Amazon will automatically drone him some L-Dopa.’ I’m willing to bet that’s already on someone’s PowerPoint at some Ted-X somewhere.” For Wilbanks, it was tempting to imagine putting webcams in his parents’ apartment. “If there’s a fall, it would trigger a ping, ‘Are you OK?’ Technically, that’s trivial. But my father wouldn’t want that. He’d hate it. On the other hand, it might liberate my mom to go to the grocery. Or see friends.” And who would pay for the webcams? The insurance company, as a way to keep people out of the hospital? Advertising companies, in exchange for bombarding the younger Wilbanks with ads for wheelchairs, incontinence medicines, burial policies? We use tech like social media to avoid things we don’t like — boredom, discomfort. “And you can imagine the thinking, ‘My mom had a fall. I drop-shipped her some flowers, and a task rabbit is building something around the toilet so she doesn’t slip anymore. Done,’” says Wilbanks. “I don’t have to call her now. I don’t have to go change her diaper.”

“Who has the right, the agency, to put in those webcams?” Wilbanks wondered. “How do you give people as they age the power to relate to others, autonomy over their aging, mastery over their health, a sense of purpose, rather than the drone dropping the L-Dopa before the doctor has even seen them?”

The way to avoid Orwellian outcomes, Topol and Wilbanks agree, is to make technology serve individuals and to make the data belong to them, rather than to the state or a bunch of corporations like Google and Facebook and Apple, which have taken possession of much of our information without really asking for it. Industry practice has been to assume that the data we provide companies is theirs to use. But many activists say that laws, or at least industrial standards, are needed to stop this from happening in the health care world — if only because the backlash could slow progress toward useful applications of technology. Too many people will opt out if it means surrendering data on their medication lists, phobias and family relations to automated bots and the individuals who profit from them, says Topol. “If we can’t get the privacy and security taken care of, this stuff isn’t going anywhere,” he said.

As an early adopter, Topol sees specific benefits from technology and shies from dystopian or utopian perspectives. But he stresses that like everything in medicine, technology needs to be tailored to the individual. “If a person has heart failure, it will be helpful to monitor vital signs from home. If there are issues of frailty, you zoom in on gait and things like that. You match the technology to the needs. You won’t want to collect data unnecessarily.”

In 2047, as today, these issues probably won’t have been resolved. Like today, we’ll be using technology more than ever, and with just as much ambivalence. Some of us, and some of our hundred million mothers and fathers, will grow accustomed to babbling with the very “smart” teddy bears that stand in for visits from grandchildren. Perhaps we’ll find it easier on our own consciences to have robots ferrying us to the bathroom, rather than inconveniencing a daughter or an immigrant nurse. Ideally, we’ll be able to use just as much technology as we need — but not more than we want — to ease the pain of growing old.