Bill Gates is naive, data is not objective

Let’s see some more example of data collection and model design not being objective:

  1. We see that cars are safer for men than women because the crash-test dummies are men.
  2. We see that cars are safer for thin people because the crash-test dummies are thin.
  3. We see drugs are safer and more effective for white people because blacks are underrepresented in clinical trials (which is a whole other story about power and data collection in itself).
  4. We see that Polaroid film used to only pick up white skin because it was optimized for white people.
  5. We see that poor people are uninformed by definition of how we take opinion polls (read the fine print).


The Bank of England Gets It ??? using search data as economic indicators

From this Bank of England paper on using search data:

???In contrast to most traditional survey methods, they are collected as a by-product of normal activity, rather than requiring individuals or firms to respond to survey questions after the event. This can avoid problems associated with non-response or inaccurate responses. And it also means

that information is continually collected on a wider range of issues, rather than just on a few pre-determined questions. As a result, search data can help analyse issues that arise unexpectedly.???

This way they can tell when the economy is pregnant…