Facts and Opinions Require Different Approaches
For 40 years, the software that businesses used took a set of rules and applied them to data. The resulting output was not much more than a reformatting of the data. Today’s tools take data and transform it into a set of strong opinions which end up as the foundation of other strong opinions.
In the first generation of software, the programmers had to understand the patterns and embed them in the code. In today’s Machine Learning era, the tools deduce and discover the patterns. We don’t actually know if this is better yet. We do know that machines are capable of ferreting out patterns where to humans there appear to be none.
This is a non-trivial change. The failure to distinguish between the two types of tool will result in significant problems. The difficulties are a part