"Voracious demand for data" of pure probabilistic reasoning

research methods design
1

Arjun Manrai has created a shiny app based on a 1978 paper by Szolovits and Pauker describing “two extremes of medical reasoning: purely categorical and purely probabilistic.” http://groups.csail.mit.edu/medg/ftp/psz/Szolovits-Pauker78/Szolovits-Pauker78.html

https://manrai.shinyapps.io/PureProbabilisticReasoning/

Comments?

2

"Although [AI] can exhibit impressive expert-like behavior, we believe that none of them is yet capable of truly expert reasoning. "

i think what it neglects is that any study data, images or whatever that you feed into the AI calculator is already imbued with so much clinical understanding, understanding of data quality etc that it can hardly be labelled ‘AI’ anymore. There is no real deduction or ‘categorical reasoning’ (i would hope) and thus always a voracious demand for data.