There are a lot of examples of these "expert systems" in medicine that seem to do well in research, but fail in other ways. We're seeing some of these failures

There are a lot of examples of these "expert systems" in medicine that seem to do well in research, but fail in other ways. We're seeing some of these failures again these days with ML models in medicine, especially bridging the gap between the proof-of-concept and production systems. https://www.jstor.org/stable/3666074 https://www.jstor.org/stable/3666074 https://aimi.stanford.edu/events/bridging-ais-proof-concept-production-gap-andrew-ng-phd https://aimi.stanford.edu/events/bridging-ais-proof-concept-...