So Appy Together

Siri, the iPhone’s sassy personal assistant, was just the beginning. Right now, behavioral scientists are racing to develop a new generation of apps and programs that can mimic compassion, concern, and sympathy—technology they hope we will form relationships with, even fall in love with. Janelle Nanos goes inside this brave new world to find out how our ever-smarter phones are changing what it means to be human.

Kvedar’s not alone. Nearly every developer I spoke with said that Siri may be a game changer when it comes to mobile health. And behavioral scientists and designers of these products are already busy building algorithms to implant into Siri’s brain. In so doing, they’re probing our psyches to find ways to get us to trust and build relationships with these tools.

Behavioral scientist Betsy Barbeau is one of these digital “neurosurgeons.” As the chief science officer at Healthrageous, the commercial arm that grew out of the Center for Connected Health, she’s designing artificial intelligence technology that will better understand us humans. People who sign up for Healthrageous are guided by a virtual coach that’s been programmed using a technique called motivational interviewing. The goal is to create an empathetic, almost therapeutic relationship between the patient and the technology. “There’s no wagging of the finger, there’s no, ‘You blew it this week,’” Barbeau explains. “It’s the idea that someone is watching you, checking in, and giving you an ‘atta girl’ if you met your goals.” That “someone” is actually a something, but who’s counting? I ask Barbeau what would happen if I lied to the coach about my calorie intake. She laughs. “Well, I could see it saying something like, ‘I see your activity level has been very high but your weight level hasn’t been changing.’ It would be an opener for further reflection.”

Knowingly cultivating an empathetic relationship with your smartphone is downright creepy. But the thing is, virtual coaches actually work. A study in the Journal of Medical Internet Research, published in January and coauthored by Kvedar and Timothy Bickmore, a computer science professor at Northeastern University, found that having a virtual coach can steer people toward better health. In the study, they provided 70 adult Bostonians with pedometers and access to a website where they could track their steps. Half of them were also able to work with an animated coach on the site that offered personalized feedback on how they were doing. At the end of the experiment, those who were coached walked an average of a half mile more per day than those who weren’t.

Bickmore says these programs are trying to replicate the communication between doctors and patients. The coach can have a caring face, he explains. “We shift its pitch and slow its voice down so it sounds more concerned. We’re trying to acknowledge the emotional dimension of what’s going on.” He hasn’t yet researched the degree to which people feel bonded to these coaches, but he’s hopeful. “A lot of time at the end of the studies,” he says, “we have participants who don’t want the study to end, and don’t want us to take the character away.”

 

Moss, the head of New Media Medicine Group, says it’s best to think of our phones not as coaches but as “partners” working alongside us. He says it’s smart to begin thinking about the emotional interactions we have with our phones now, before we get to the point where we’re breaking up with devices or wondering why they aren’t returning our calls. But he also cautions that the fear that we’ll become cyborgs shouldn’t deter us from charging full bore into the digital health space. For one thing, he argues, the more data we collect about ourselves, the more it can be used to help feed into the population, giving us a far more accurate way to shape public health policy.

I ask Sherry Turkle, the MIT professor, if she can foresee a day when she views her phone as a partner.

She pauses. “I get migraines,” she says — the close-the-drapes kind that keep her in bed for days. If a phone had the capacity to measure her temperature, pupil dilation, and galvanic skin response — three things that indicate the onset of a migraine — and then tell her to take a pill to fend off the massive headache, she would absolutely want it. It would be akin to having a hero rescue a damsel from the path of an oncoming train.

When I ask Turkle how she’d feel about a device that could do all that for her, she lets out a laugh.“Could youimagine the feelings of intense connection and gratitude, bordering on profound love?” she says. “It would be almost impossible not to love them.”

At which point, I get the chills. You see, it makes me anxious to think about it, but the truth is, I know I want that, too. The allure of knowing so much about myself — even if it means being inextricably linked to a tiny gadget in my pocket — is too strong. When the time comes, I’m going to surrender.

Turkle says that as the technology advances, we’re going to have to figure out how to deal with such a reality. “We have to learn how to exert a certain kind of discipline to remember that these devices don’t care about us. That they don’t love us. That they’re not looking out for us. That there’s nobody home.”

Welcome to the next stage of human evolution: playing out in the palm of our hands.

ADVERTISMENT

  • Andrew

    One thing I worry about is that these apps can further exaggerate the relevance of surrogate or intermediate outcomes to our health, so that we are pushed to produce the best lab test results, instead of living the most healthful lives that fit with our goals and values.

    Too many patients are already obsessed with the test. For example, a survey of people with diabetes found that one in four considered their hemoglobin A1c levels to be a more important outcome than even death, stroke or heart attack!
    (Murad MH, Shah ND, Houten HKV, et al. Individuals with diabetes preferred that future trials use patient-important outcomes and provide pragmatic inferences. J Clin Epidemiol. Jul 2011(7):743-748.)

    Such distortions occur when the first thing clinicians talk to their patients about is their test results. Patients know they’ll get gold stars for good lab tests and a harrumph for missing the mark.

    Sure, people with diabetes should be aware of hemoglobin A1c, heart attack survivors should track their LDL, and so on. But we should not let these numbers consume our attention far out of proportion with their actual usefulness to promoting health.

    Smartphone apps are great at tracking numbers, but can any track how many times we…