Author of the 2017 book MoneyBall Medicine and host of his successful eponymous podcast series, (which is globally ranked in the top 3% according to Listen Notes), Harry Glorikian, a seasoned life science executive and investor took some time out to talk to Damian Doherty, IPM’s Editor in Chief about his new book.
Q: What was the inspiration to write the book?
A: I’ve been involved in healthcare technology for several decades and am a self-professed tech junkie. I did my best to capture some of the rapid changes in my last book – MoneyBall Medicine. But that was written for someone working in the industry of healthcare/life sciences. Then a few years ago I was pleasantly surprised when my Apple Watch and an app called Cardiogram analyzed some data patterns and asked me if I had been diagnosed with sleep apnea—something I had recently been diagnosed with. I know that some people might feel uncomfortable with their wearable devices like a smart watch or fitbit figuring out that you might have a health condition, but I felt the complete opposite and was intrigued to learn just how much more these technologies could do. This whole concept of gathering and leveraging data for every individual is changing everything, and I wanted to share how it is and will have a profound effect on helping people improve their health. I think if people understood and appreciated the impact these technologies can have on their well-being, then we could completely change the how people manage their own healthcare.
Q: Is a third book easier to write and did AI help you in any way?
A: My actual writing process from book to book hasn’t changed. Between the discussions I have on my podcast and all the existing ideas bouncing around my head. My family and friends get a barrage of “Hey, did you know….” and “Have you heard about…” as I start to home in on the topics I want to focus on. It’s an iterative process of drafts, taking a small break before going back to a chapter with fresh eyes. And with this book, I really wanted to make sure I was making the language accessible to a non-science audience, so I asked my wife (who was an English major) to read everything at least once, too.
Regarding AI helping the simple answer is I wish but not yet. Maybe someday in the near future.
Q: Did you have an audience in mind when you started this project.
A: The Future You is written for everyone. Truly. From your tech-savvy teenager to your aging tech-adverse parent. I had to take a step back, think how I would explain this concept to my neighbor or my aunt. It wasn’t enough to write about topics that Harry is already interested in, it was how do I find the topics that have broader appeal and then how can I write about them in a way that gets others as intrigued as I am. I’ve done my best to strip out the medical and technical jargon to explain key concepts in a way that makes sense to everyone who doesn’t have a medical degree. And for doctors, scientists, others in the healthcare or research field, there’s plenty of fascinating examples of how technologies like artificial intelligence and genomic testing are quickly changing how things are done. I really think there is something for practically everyone in this book. And even though I made the language more accessible, I also made sure to include hundreds of references and links to websites at the back of the book if you want to do a deeper dive on your own.
Q: Your book articulates comprehensively how AI is stitched into the fabric of our everyday lives and its power to effect positive change, particularly in medicine, is irrefutable. What areas concern you most around AI and it’s potential misuse?
A: The three areas that concern me are regulation, policy, and security. Regulation needs to be set properly to ensure that these devices are performing at the right level and giving the user information that is meaningful to their condition. Policy is an area that concerns me because it is woefully behind the exponential growth of where technology is today and where it is going. Finally security, we need to make sure these systems are as secure as our banking systems.
Q: Do you think we’re doing enough to eliminate bias and protect the integrity of the data being fed into AI systems?
A: The short answer is yes. I believe one approach is algorithmic hygiene. Making sure as much as possible that training data are representative. No data set can represent the entire universe of options. Your need to try to identify the target application and audience upfront, and then tailor the training data to that target. There are many other approaches that can be taken but probably for a longer discussion at some point. The one thing to know is there are multiple approaches to eliminate bias in AI, and none are foolproof. We need the field of ethical AI to develop quickly and identify different approaches that can be used for highly regulated fields such as medicine, where transparency and explainable AI are of critical importance.
Q: Rewind the clock back to your life as a teenager which app would you have been useful to you back then?
A: Probably the workout apps since I was very active and always trying to optimize my performance. But I do wonder if I would have appreciated the technology and what it was telling me. The problem is when you are young you feel invincible.
Q: Do you think AI is the panacea for healthcare’s ills?
A: I truly believe that technology + the trained physician is the ideal combination for optimal or better outcomes. The technology can give the physician longitudinal data and/or insight that allows them to manage their patients better.
For the individual the technology acts as a dashboard on their personal health/wellness. This give them the opportunity to course correct along the way which hopefully results in avoiding certain chronic conditions that happen because you were not empowered with the data.
Q: When an AI assisted decision proves fatal, who is responsible? The company that developed the AI or the physician that used the AI?
A: Loaded question. I believe this needs to be reviewed on a case-by-case situation with the need to understand the details. I believe we are asking similar questions when it comes to autonomous cars. I believe many of these questions will be worked out over the next 3-5 years.
You talk about CRISPR’s huge potential in enabling precision medicine in your book. Can you envisage another Nobel Prize winning discovery like this in the next decade?
Yes, I do believe this is possible and even probable especially with the assistance of AI and the tool box it provides us. The ability to identify new discoveries utilizing this toolbox will something to watch over the next 3-5 years.
Q: If you could devise one application for healthcare that your children will benefit from in their adult life but that would be deemed science fiction today, what would that be?
A: Well one of them is already being developed and that is liquid biopsy. Being able to detect cancer early will be a benefit to everyone. If I had to pick a second one it would be a wearable or app that could look at multiple biomarkers simultaneously to be an enhanced dashboard for optimizing health/wellness.
Q: Do we need to be thinking of the educational curriculum and developing a new breed of data doctors to support this new world of AI?
A: Absolutely. But I would give you the same answer for genomics.
I believe the medical schools need to implement some dramatic changes in the curriculum to keep up with the pace of change taking place. The current system is not fair to doctors in how it prepares them for what is here now and what is coming in the next 3-5 years.
Q: We’re all trying to simplify our lives and AI is certainly a tool that can save the precious commodity of time. Conversely a lot of these apps can feel onerous and overwhelming. How do you strike the balance?
A: My hope is that the companies developing these products devise better ways to communicate the data to the user making it easier to understand and therefore decreasing the burden on the individual.
But I always tell people who are trying to incorporate something new into their lives to choose one thing and make the most out of it before introducing a new device and/or app.
Q: We’re living in turbulent times and despite being more connected in many ways through tech we’re seemingly very disconnected and divided as a society. Do you think AI has the power to transform humanity for the good?
A: This is a concern of mine and a question that requires time to answer. That said – Technology is moving faster than humans can adapt and the creators of this technology cannot fully appreciate what the implications of it are. Therefore, I always encourage policy makers to do their best to stay ahead of the curve rather than behind it.
Q: You talk about the advancements of genome sequencing and indeed pandemic preparedness in the book. Global genomic surveillance has played a significant part in monitoring the evolution of the virus and it will be critical to how we deal with the next outbreak. Would you agree that multiple countries will need to invest heavily to build the infrastructure rapidly – otherwise we’re going to be left facing many of the same challenges again?
A: Yes absolutely. Pandemics need global surveillance. We need sequencing on a global basis with the ability to exchange/centralize the data. We need the systems to be as fast as possible from sample to answer and we need advanced predictive analytics to help decipher and present results.
Q: Imagine a hero of yours, past or present, has emailed you and said – ‘Hey Harry, let’s do lunch. I just read your book and I loved it but I have some questions for you!’ Who would you like that to be?
A: Interesting question – there are so many great people. If from the past I would like to sit with Benjamin Franklin. If from the present I think that would be Elon Musk.