“Healthcare is the biggest business in the world, and it is phenomenally broken,” says Peter Diamandis, cofounder of the X-Prize, Singularity University, and Health Longevity Inc. “So, do I think Apple and Google and Amazon can do a better job? A thousandfold.”
In his upcoming book, The Future Is Faster Than You Think, which will hit bookshelves in late January 2020, Diamandis makes the case for why he believes big tech companies are going to be running healthcare by 2030. In December, he came to Fast Company’s offices to make the case for why Big Tech is the doctor of the future.
“We’re going to see Apple and Amazon and Google and all the data-driven companies that are in our homes right now become our healthcare providers,” he says, referring to smart speakers such as Google’s Assistant, Amazon’s Alexa, and Apple’s HomePod. While many of these home voice assistants started with simple tasks like restocking home pantries and surfacing cooking tutorials, they’re already starting to move into the business of managing family well-being.
Amazon has put significant effort into making Alexa a health resource. In the United Kingdom, it has partnered with the National Health Service to answer basic health questions such as “What are the symptoms for shingles?” or “What do you do if you have a cold?” It has also made Alexa compliant with U.S. HIPAA laws and signed partnerships with major healthcare insurers and providers so patients can access or remit health information through the device. To date, there are nearly 2,000 health wellness skills on its platform.
Healthcare is the biggest business in the world, and it is phenomenally broken.”
Similarly, the Google Assistant uses search to serve up information about medications, symptoms, and diseases, as well as physicians and medical services. Both the Google Home and the Echo have a Mayo Clinic-developed skill called First Aid that helps people navigate minor injuries. Meanwhile, Apple’s HealthKit takes a slightly different approach to tackling personal health. The kit connects to Apple’s own products such as the HomePod, iPhone, and Apple Watch as well as a bevy of devices from other companies, such as scales and blood pressure cuffs. The HealthKit can also tap into electronic medical records and other apps connected to hospitals and doctors. Essentially, it becomes a single repository for all your precious health data.
Diamandis believes the involvement of home health devices has the potential to lower costs by shifting care away from hospitals, where expenses can be much higher. This is the general idea behind telemedicine, but Diamandis thinks that big consumer tech companies will play a big role in driving that vision. He also thinks that these companies, which have mastered using personal data to anticipate user behavior, can use personal health data to make predictions about a person’s long-term health prospects and advise them accordingly.
Diamandis posits that the more information is available about you—your genetic makeup, your health history, what you ate for breakfast, the bacteria in your bowel movement, how you slept last night, what kind of sound you’re exposed to every day—the better artificial intelligence will be at spotting your potential for illness and suggesting care before the problem becomes intractable. This approach might shift the medical establishment from a structure that treats disease once it’s wreaking havoc in your body to one that prevents the disease from striking in the first place. “It is literally hundreds if not thousands of times cheaper to do that,” he says.
It is literally hundreds if not thousands of times cheaper to do that.”
It is this cost savings that he believes will allow for new models of healthcare. Diamandis predicts Apple and Amazon will come up with a service where a person pays a company to keep them healthy, rather than to cover the cost of illness, based on their health history and daily activities. And big tech could not only influence a person to make healthier decisions, it could force them. Amy Webb, professor of strategic foresight at New York University’s Stern School of Business, has spoken at length about the possibility that in a futuristic situation when Amazon, Google, and Apple run your entire house as well as your healthcare, smart refrigerators could cut you off from snacking between meals and smart garages could keep you from accessing your car in favor of walking to work.
Diamandis believes that by knowing a person’s predisposition for disease, these companies could help them live a healthy lifestyle with their particular abnormalities in mind. “Can you prevent those things, so we don’t have these extraordinary costs?” he asks. It will be these services, he believes, that will lead healthy people to dispense with traditional health insurance, leading to its ultimate demise.
The problem with precision medicine
Diamandis’s vision of healthcare in 2030 raises a lot of questions. First and foremost, do these big tech companies want to become healthcare providers? So far, the only one that has really signaled its desire to become your doctor is Amazon. In addition to its work with Alexa, the company has launched its own health clinic for employees and is working on a secretive health project with JP Morgan and Berkshire Hathaway called Haven. But Apple and Google, at least so far, seem content to integrate their technology with traditional health providers as a way of advancing their practices. Meanwhile, the insurance industry is more likely to adapt to a preventative health model than it is to collapse completely. A survey from last year shows insurers are increasingly signing contracts with healthcare providers for continuous, value-based care—all for a flat rate—rather than a negotiated fee for a particular service.
But Diamandis is right to bet on artificial intelligence in some regards; it is already predicting the onset of disease with some success. What’s unclear is how far forward these predictions can reach and how meaningful big data is to understanding how our bodies work. For example, while it may seem clever to sequence the genome of every new child born, one of Diamandis’s ideas, it actually isn’t as effective as a blood test for catching certain disorders, reporting has shown. Furthermore, the promise of predictive medicine may rest on a flawed assumption.
In a recent paper, Henrik Vogt, a post-doctoral fellow at the University of Oslo Center for Medical Ethics, lays out why big data may not deliver in the way Diamandis suggests. He says that as technology gets better at spotting indications of illness or the prospect of sickness in the body, it will surface more and more signals. But a predisposition for a disease does not equal a diagnosis. “The main problem for big data screening is that monitoring many features of the body with highly sensitive technologies is bound to detect many abnormalities but without the ability to tell which, if any, will become clinically manifest. As a result, more people may be labeled with more harmless conditions,” he writes.
We have to accept that there will always be some degree of risk, morbidity, and mortality.”
Even if a person has a high likelihood for a disease, they may never present symptoms, Vogt notes. As more services and devices—such as direct-to-consumer gene sequencing and wearables with heart rate variation detection—get more sophisticated, there is more visibility into a person’s body. But there is also a lot of noise in this information. Not every little genetic abnormality may be meaningful. Different bodies may have different idiosyncrasies. While there is more room for prevention as we are all more aware of our disease risk, Vogt makes the case that there is also a risk of overtreatment, which could be costly and may also cause patients harm. Vogt also explained via email that there might be issues in investing too much in big data rather than another approach, such as social or institutional change.
That is not to say there isn’t a huge opportunity to mitigate disease through data and intelligence, Vogt writes, but doctors need to rethink risk. “We have to accept that there will always be some degree of risk, morbidity, and mortality,” Vogt writes.
That perspective flies in the face of precision medicine, which tends to assumes the human body is like a machine, Vogt explains over email, something that can be measured, analyzed, and ultimately controlled. “The historian Yuval Harari, for example, rather uncritically built his book Homo Deus on this assumption: that ‘organism is algorithm,’” he says. But human bodies don’t work like that; they are unique in composition and environmental circumstance. “Both for biological and statistical reasons, there are limits to how precisely and accurately the trajectory of a human life can be predicted. This obviously limits the promise of predictive medicine.”
This point of view is crucial, because it is at the heart of some of the skepticism surrounding a big data-focused approach to medicine. It is the reason that Apple has doctors on staff to advise on the development of its medically minded hardware. For big data to really drive better health outcomes, as Vogt points out, there will have to be standards about what information is actionable and what is not.
Diamandis seems to concede that big data is not everything, “Ultimately what’s best is human and AI collaboratively,” he says. “But I think for reading x-rays, MRIs, CT scans, genome data, and so forth, that once we put human ego aside, machine learning is a much better way to do that.”