Researchers and technologists spend far too much time focusing on the sexy what-might-be, and far too little time on the important what-is.
to address the problems that AI is creating now, we need to understand the data and algorithms we are already using for more mundane purposes.
