AI, Health, And The Future Of Human Agency

Next Now: The 21st century has ushered in a new age where all aspects of our lives are impacted by technology. How will humanity anticipate, mitigate, and manage the consequences of AI, robots, quantum computing and more? How do we ensure tech works for the good of all? This Ashoka series sheds light on the wisdom and ideas of leaders in the field.

Dr. Stephen Friend is a globally acclaimed serial entrepreneur and biomedical researcher. He currently is the President and co-founder of 4YouandMe, a visiting Professor of Connected Medicine at Oxford University and Chairman of the Board of Directors of Sage Bionetworks, where he was co-founder and President. He previously held positions at Apple, was Senior Vice President at Merck & Co, founded and led Rosetta Impharmatics, and was an Associate Professor at Harvard Medical School. He was elected an Ashoka Fellow in 2011. Before venturing into medicine and cancer research, he studied anthropology and philosophy.

Konstanze Frischen (Ashoka): Stephen, how can AI help us live healthier?

Stephen Friend: It’s simple: Wearable devices – rings or watches that register heartbeat and all kinds of other physiological data – today can allow individuals to follow themselves. But the potential is bigger: What we are betting on is that we can use AI and machine learning to analyze these data in ways that better track stress and individual symptoms so we can forecast the onset or worsening of chronic diseases like migraines, MS, Crohn’s disease, or even vulnerable times of metamorphosis, like pregnancy or menopause.

KF: And what are you learning?

SF: We are learning how to collect multidimensional, longitudinal data on chronic illnesses – and the results are remarkably strong. An individual’s physiological data – variable heart rate, breathing patterns, skin tone, and much more – yield patterns that correlate with flares of symptoms. And we are setting up studies to test if the more centered a person is in themselves, the lesser the degree of their illness. Conversely, we can see that stress is a breeding ground for symptoms.

KF: So, the data generated by wearables show that our states of mind like stress or wellbeing are underpinned by physiological markers that correspond to diseases?

SF: Yes, this is what we expect to verify. The signals are numerous, and in unison. The body has a way of responding to stress that changes your voice, blood pressure, sleep, what’s going on in your pancreas, your sweat and so on. And the ability to pick these data sets up and stick all of them together, means we can get a pretty good idea of how someone is doing and can start to make individual forecasts for a patient when a disease is likely to start or flare up.

KF: So, AI generates data that allows us to predict our health. But what happens then? Who makes the decision what to do next?

SF: This is the key ethical question. It could go both ways. A central health AI system could prescribe across a whole population what it considers to be the maximal “public good” to combat, say, obesity. As in: person #1, Stop eating carbs, person #7, Start exercising. Or as an alternate, we could give individuals a dashboard with their own data and ask them: What do you want to do? I’m concerned that we might lean towards the former option. Why? Because the machine will likely be good at making the decision for you. Extraordinarily good. So much so, that from an efficiency standpoint  – let’s imagine a government that wants to bring down obesity by 30% – it would be easier for that government or healthcare executive to instruct machines: give these directions to those people at these times.

KF: That’s scary.

SF: Yes, very. What is it going to feel like to be a person in a world where decisions about your health get made for you and imposed on you? Probably not very good. So, the question is: What does it take to ensure that everyone has agency? In the new era, how can we make sure people are in charge of their own health conditions? That is what we are working on in our projects.

KF: How do you do this?

SF: Our hope is that if we give patients the right knowledge about their health and the right emotional and social support structure, they will be able to have agency: they can make informed decisions about their life, and get the support they need to do what is best for them. People will need to trust each other, to be willing to share their symptom data and their insights and together build a set of supports and knowledge, by each other for each other.

KF: That sounds very different from the experience I have currently when I see a doctor.

SF: Completely. Currently, your doctor with a deep desire to help is likely functioning as a somewhat out of date parochial AI machine – relying on their brain and sources of knowledge close at hand. They’re trying to make decisions in your interest, but can’t possibly have a deep knowledge based on the vast complexity of data that now exists and will soon explode. They look at your symptoms, and then make decisions not unreasonably off the things they learned in medical school, their experiences in practice, or their refresher courses. I do not want to sound harsh, but it is worth remembering that our chronic care today is administered by anointed guilds of experts using orders set down by regulatory agencies, primarily guided by often moderately relevant studies.

KF: You went to medical school yourself, but your view of the medical profession seems to be, let’s say, less than high.

SF: It’s really strange for me to have people saying I don’t know what it would be like for an AI machine to replace my doctor’s decision making – with this underlying concern that the health AI expert machine might have biases. Before getting too concerned about the AI expert biases, let’s remember the range of mild to atrocious biases that are in all of us, be we doctors, reporters, or craftsmen.

KF: So, what is the future going to look like?

SF: Critical health support will increasingly come from the patient’s community – from family, neighbors, nurses, social workers. A good example today is the approach being used by 7 Cups of Tea to help those with mental health issues. Wearables will be able to tell patients with chronic illnesses that if certain symptoms are detected, an oncoming spike might occur, and maybe even alert us to things we want to do to avoid. Wearables are likely to give us nudges such as: Have you considered how you might lower your stress levels or obtain emotional support. And then, the community will provide the interpersonal support needed to help the patient with their choices. It is already possible to foresee a time when a patient will only infrequently need a doctor or a life coach with an outrageous salary.

At the same time there is an evolving role for the clinician, but it will be as an integrator to help improve the intelligence of the master system and to help people nourish themselves at times of metamorphosis. Please remember – I am talking about a future state where AI will augment our health decisions. As this is out there a decade or more, the next years will be key ones: medical experts and engaged participants together can reshape chronic health care. This is a partnership where the AI tools will become better at giving people what they desire- ways to track their own health trajectories as compared to others, in ways that enable wisdom and support from those with similar journeys.

KF: You’re painting a picture of a radical new model of health: AI backed, community based, centered around a patient’s agency. How do we get there from where we are now?

SF: I’ll give you an example of what a pathway can look like: Our nonprofit 4YouandMe is about to start a study on menopause with wearables first highlighted by Prof Becky Malby at LSBU in London. The study is currently proposed as one to be owned and led by the participating women. In this format they will set the questions, what they want to learn, and they will support each other on their journeys. It could hopefully become a clinical group co-operative community. Ultimately, though, we will need a new architecture and a new economic structure. I just don’t see that coming from the commercial interests currently working in AI and health.

KF: Why not?

SF: There is a fundamental disconnect between the capitalist structure of the current medical system and the benefit to the individual. It’s like in the energy industry. For an energy supplier, their interval of profit-making is such that they are not incentivized to look at the environmental consequences. It’s the same in the healthcare system. And with companies making wearables – they are going to look for optimizing that 30% weight reduction we spoke about earlier, but I don’t think they are going to look out for the wellbeing of an individual, and they are not likely to jump-start the revolution we need to provide each individual the knowledge and support to better self-navigate through times of metamorphosis.

KF: How do you think can we change that?

SF: Those who are in the pathbreaking position are nonprofits, social entrepreneurs, universities and patient advocacy groups. They can model a new way that advocates for the good of the individual and their agency. Ultimately, our new model will have economic benefits for society as a whole because we will dramatically lower our health care costs. But currently, we are caught up in this acute heavy, late stage heavy, treatment and heavy way of healthcare especially in the U.S. A doctor is interested in giving you a drug, a treatment, or a referral – as long as it involves a financial transaction. But with the nascent abilities of our smartphones and wearables, we have the opportunity to collect and give back the information each of us generates, to support and nourish the return of agency to individuals. We have the chance to shift and prevent chronic illness a significant portion of the time without incurring much of the huge costs built into the current system of medical care.

KF: So, there is a huge potential for AI and the wearable industry to align with social organizations to re-vamp the health care field.

SF: Yes. The fun is that AI and wearables are poised to enable us to crack open the sacred roles in the field of medicine, and to give agency back to the individual. But because our new model of health care will not incur the same costs, hence not the same profits, there will be less interest from companies at least at the beginning of this fundamental transition in health care. Advocacy and leadership will come from the nonprofit side.

Konstanze Frischen leads Ashoka in North America and the network’s emerging focus on AI, technology, and ethics. A journalist and entrepreneur, she previously founded Ashoka in her native Germany.

The article was origianlly posted at: %xml_tags[post_author]% %author_name% Source%post_title%