A couple months back we made our first big data investment in the healthcare space. Iodine was founded by a stellar team that includes Thomas Goetz (former Executive Editor at Wired and author of The Decision Tree) and Matt Mohebbi (co-founder of Google Flu Trends, perhaps the world's best known application of big data in health today). Thomas conceptualized the company out of a TED talk he gave in 2010 and our first diligence call with him was highly stimulating. One of the thoughts that came out of the chat seemed particularly relevant to our cross-border investing thesis at Amasia.

Big Data and Traditional Chinese Medicine have a lot in common. 

In the sparse data world of yesterday, scholars would think up some hypothesis and design an experiment collecting a limited number of factors to test this hypothesis utilizing complex models to compensate for paucity in data. This very much mirrors the development process for pharmaceuticals, which depend largely on clinical trials with little understanding of how drugs perform in a messy chaotic world. 

In the big data world of tomorrow, cheaper sensors, faster connectivity and broader internet penetration are all converging to empower us to measure everything all the time. The datasets are infinitely larger and significantly messier, but when analyzed properly can yield insights that scholars would never have hypothesized and tested in a controlled setting. Our process for collecting knowledge is rapidly becoming defined by correlation and not causation. 

My mother once told me that TCM developed at the riverbeds in ancient China, where droves of villagers would congregate to do their laundry and exchange information on anything and everything. Though their analysis did not require hadoop, their data was messier than that of most FDA studies, and their sample sizes certainly larger. TCM does have some overarching frameworks defining the flow of chi or energy through the body, but the development of treatments resulted from a correlation-based approach rather than a causation-based one. 

(As a sidenote, my mother has her PhD from MIT in analytical chemistry and is generally a wise woman, but certainly not an expert in the history of eastern medicine. So please take this with a large pinch of ginseng). 

As we move from the riverbed to the internet, from datasets of millions to billions, companies like Iodine promise to launch a correlation based approach to healthcare into an entirely new dimension. We can't wait to see what happens next.

 

Posted
AuthorJohn Kim