“120 over 70…good”. I nod in mock comprehension as the nurse reads my blood pressure from the chart. I think about my blood pressure precisely once a year, during an annual biometric screening. Since I’m in relatively good health, the ritual reading of the two numbers is usually followed by a perfunctory positive comment - providing me with about one second’s worth of satisfaction before completely disappearing from my consciousness until the next year. Ultimately, those two numbers don’t factor in to my day-to-day decision making at all.

And yet, we must make decisions that affect our health every day. Whether choosing what to eat, when to exercise, or how to treat our chronic back pain, we often decide based on little more than a hunch or a cursory Google search. And we make most of these decisions in isolation, not as part of a larger plan. Think of this popular approach as the Hope For the Best™ Strategy, which only serves to ingrain whatever default choices we find in our environment.

Asking for help seems like a good idea in the abstract, but we typically don’t seek professional advice until something has gone wrong. Doctors orient their services around sickness, not unleashing your most kick-ass self. Their understanding of each patient is like a blurry picture, developed during the examination (and maybe a few tests). Again, health systems are oriented around fixing problems, not optimizing against your health goals.

Staying healthy is up to you. Unfortunately, many of the questions you’ll face don’t have clear answers, even for the pros. For example, will spinal fusion surgery reduce your chronic back pain more than physical therapy? The evidence doesn’t support it but there are still plenty of surgeries done every year (on the order of hundreds of thousands). The more nuanced answer, of course, is that the right intervention depends on your goals and circumstances. And let’s be honest, you’re not going to review the health literature or call a doctor each time you have a twinge of pain or sit down for lunch.

It gets worse (sorry). Even if you had a doctor following you around all day, you would still routinely make suboptimal decisions. Decision making in Health is complex, and it shares certain inherent problems with decision making in other complex, dynamic systems. I’ve broken down a few of the issues to explain what I mean:

1. The Aristotle Problem: Every expert is constrained by the limits of their knowledge. Even if you had a genius of historic proportions (like Aristotle) as your 24/7 personal assistant, you would still be limited by what they know.
2. The Librarian Problem: There is too much available knowledge in a given domain for anyone to sift through it all in a reasonable amount of time. Libraries are useful as content repositories, but it is the librarian who saves you time by directing you where to look next.
3. The Superman Problem: It is difficult to translate intentions to actions that will have outsized effects (i.e. leverage points). How does Superman know who to save next?

Let’s pause for a moment to reflect on how cool it would be to have a personal cabinet made up of Aristotle, Superman, and a really good research librarian. Recognizing that it’s probably never going to happen, how do we move past these limitations? What we need is a system that is unbounded in its knowledge and perspectives, relentless in its pursuit of answers, and able to optimize against particular goals.

I see hope in the Quantified Self Movement and companies like Google, IBM, 23andMe, WellnessFX, Basis, and Lift, among many others. There is a rising awareness that technology can address specific, personalized health questions. However, we’re still left with a huge number of inputs into the decision-making process. I can’t be expected to navigate all of these services and combine them into an useful mix.

What we need is an advisor layer that acts as an intermediary between a single user and the underlying recommendations. It would sit on top of existing services, surveying all known outcomes in order to build optimal, personalized paths for any user. This approach will combine the codified learning of the best experts, massive amounts of data, and human judgment.

Honestly, I’m not sure exactly how it would work, but I’m excited to help build that future. Until then, I still have to figure out what to do with these two numbers.