Bits of Information


The Production and Application of Knowledge

Wired had an interesting article about how Watson, the famous IBM computer that defeated humans at Jeopardy, is better than doctors at diagnosing cancer in patients. I quote the most interesting paragraph for my purposes here:

According to Sloan-Kettering, only around 20 percent of the knowledge that human doctors use when diagnosing patients and deciding on treatments relies on trial-based evidence. It would take at least 160 hours of reading a week just to keep up with new medical knowledge as it’s published, let alone consider its relevance or apply it practically. Watson’s ability to absorb this information faster than any human should, in theory, fix a flaw in the current healthcare model. Wellpoint’s Samuel Nessbaum has claimed that, in tests, Watson’s successful diagnosis rate for lung cancer is 90 percent, compared to 50 percent for human doctors.

First some initial thoughts: the idea of needing 160 hours per week to keep up with what’s new in your field is an incredible fact. There is no way to hope to be on top of everything. It is no wonder that doctors use relatively little evidence based medicine in their practice. There is simply too much evidence, which implies a high degree of information or knowledge congestion.

Second, the differential in success rates in predicting cancer between doctors and Watson is absolutely astounding! Doctors are only right about half the time (no better than a coin flip). Enter Watson, who gets a diagnosis right 9 times out of 10. This is as clear an opportunity as I have ever seen at improving the efficiency any process.

So, what are some implications of this? 

  1. We, as individuals, cannot keep up with our own rate of knowledge production in any field. So if we want to get really good at something, it has to be increasingly narrow in scope (Ben Jones at Northwestern calls this the ‘burden of knowledge’).
  2. Computers are proving to be increasingly more adept at the application of knowledge (hence all the excitement over big data).

Taken together, it seems that people ought to focus on their comparative advantage of producing new knowledge and we should increasingly rely on computers to process and apply that information. Of course, both activities inform one another, so this is likely to be a cyclical relationship.